Condition monitoring: 5 technologies to prevent machine failures

Table of contents

Condition based monitoring: 5 technologies to prevent machine failures

In industries where machines power critical processes, even minor breakdowns can lead to significant costs. That’s why condition monitoring is essential. It helps detect developing issues early, preventing downtime, reducing repair costs, and extending the life of your equipment.

This blog will explore condition monitoring technologies, helping you choose the right solution for your needs. We’ll dive into five key data sources and explain how they can protect your machinery.

Why condition based monitoring is important

Machines in industries like food, water, fuel, and manufacturing often face tough conditions. Over time, wear and tear is inevitable, leading to breakdowns. Condition monitoring identifies faults before they become serious, allowing you to address issues when repairs are cheaper and less time-consuming.

Beyond preventing machine failure, condition monitoring can also:

  • Improve safety by reducing the risk of dangerous breakdowns.
  • Increase efficiency by providing insights into energy use and system performance.
  • Promote sustainability by reducing waste and prolonging equipment life.

Condition monitoring is especially useful for critical assets: machines where failure would have significant financial, reputational, or safety consequences.

Five key condition monitoring technologies

Condition monitoring relies on data to track machine health. There are five major types of data sources used for this purpose: electrical signals, lubricants, stress waves, temperature, and vibrations. Each technology has its strengths, depending on the type of machine and potential failure mode.

1. Electrical Signature Analysis

Electrical Signature Analysis (ESA) is one of the most powerful techniques in condition monitoring. ESA works by analyzing the current and voltage of a running machine, using sensors placed between the power supply and the equipment. This allows you to continuously monitor equipment without stopping operations.

ESA can detect developing issues by analyzing changes in the electrical signals, such as:

  • Voltage unbalance
  • Harmonic distortion
  • Lateral and torsional vibrations

By continuously monitoring these signals, ESA offers real-time insight into the machine’s health. This technology is particularly useful for motors, generators, and other rotating equipment.

2. Lubricant Analysis

Lubricants play a critical role in maintaining bearings, gearboxes, and hydraulic systems. Lubricant condition monitoring techniques, such as atomic emission spectroscopy or Fourier transform infrared spectroscopy, can detect signs of wear, contamination, or aging.

Lubricant analysis can provide early warnings of:

  • Contamination by water or foreign materials
  • Metal wear, which is detected through trace amounts of iron, copper, or lead in the lubricant
  • Oxidation, indicating that the oil is aging and losing its effectiveness

Historically, lubricant analysis has been done offline by collecting samples. However, real-time sensors now allow for continuous monitoring of lubricants. This means you can detect issues like oil degradation and contamination immediately, avoiding further damage.

3. The role of stress waves in detecting damage

Stress wave analysis, also known as acoustic emission, measures elastic waves within materials. When cracks, leaks, or fiber breakage occur, they generate stress waves, which sensors can detect.

Stress wave monitoring is highly effective for detecting:

  • Surface cracks in metals, concrete, and fiberglass
  • Friction and wear in machine parts
  • Corrosion fatigue in high-stress environments

This technology is often used to locate the exact source of a developing fault, such as cracks or pitting. However, it’s important to note that stress wave sensors typically need to be placed close to the affected area, meaning multiple sensors may be required for large machines.

4. Temperature monitoring for early detection

Temperature monitoring is one of the most straightforward condition monitoring techniques. By using thermocouples or infrared cameras, temperature sensors can detect overheating in motors, bearings, and other components.

Examples of what temperature monitoring can identify include:

  • Overheating bearings, which may indicate under- or over-greasing
  • Thermal damage to stator windings caused by voltage imbalance or ventilation problems
  • High temperatures that signal impending equipment failure

While useful, temperature monitoring often detects issues after they’ve already begun causing damage. It’s not typically used as an early warning system, but it can help track equipment health in real time.

5. Vibration monitoring

Vibration monitoring is perhaps the most well-known condition monitoring technique. Vibration sensors measure the oscillating movements of a machine, which can signal issues such as misalignment, imbalance, or bearing wear.

Vibration analysis can detect:

  • Misaligned couplings
  • Gearbox failures
  • Pump cavitation, caused by bubbles forming and collapsing in liquids

Vibration sensors can be placed directly on the machine to monitor displacement, velocity, or acceleration. With continuous monitoring, vibration analysis offers early detection of developing faults, making it a key technology for critical assets.

Matching the right technology to your needs

Choosing the right condition monitoring technology depends on the machine and the type of failures it is prone to. For example:

  • For bearing wear, vibration and lubricant analysis are effective early indicators.
  • To detect stator winding issues, electrical sensors provide real-time insights.
  • In cases of pump cavitation, vibration and electrical sensors work best.

The P-F curve (Potential-Failure to Functional-Failure curve) can help guide your decision by mapping how early each technology can detect faults. The closer a technology detects the issue to the potential failure (P) point, the more time you have to address the problem before functional failure (F) occurs.

Fmeca helps you qualify and quantify the impact and likelihood of machine failure.

FMECA helps you qualify and quantify the impact and likelihood of machine failure.

Sample p-f curve for a specific failure in a specific production system. The locations of the various signals on the curve will be different for each machine, production environment and failure mode.

Sample P-F curve for a specific failure in a specific production system. The locations of the various signals on the curve will be different for each machine, production environment and failure mode.

Example 1: Detecting bearing failure early

Bearings are critical components in many machines, and their failure can lead to significant downtime. Vibration monitoring is one of the best techniques for detecting bearing damage early. Over decades of analysis, vibration monitoring has proven highly effective, offering a well-established set of standards and known patterns to identify issues.

Vibration sensors can detect several early indicators of bearing failure, such as:

  • Increased vibration due to bearing wear or misalignment.
  • Unusual oscillations, which may point to insufficient lubrication or contamination.

Another common cause of bearing failure is lubrication issues, such as water contamination or lack of proper lubrication. Lubricant sensors are the best choice for detecting contamination by fluids or foreign particles. They can catch this problem early by monitoring the chemical composition of the lubricant, such as the presence of metals, which indicates wear.

Temperature sensors can also play a role, but they tend to detect problems after they’ve started to cause damage. Overheating bearings may signal over-greasing or under-greasing, but temperature increases often occur after damage has been done. Therefore, temperature is more of a late-stage indicator.

For bearings within motors, electrical sensors can also detect issues, particularly bearing currents caused by variable frequency drives (VFDs). Electrical monitoring can catch these currents early, preventing overheating and further wear in motor bearings.

Quick-and-dirty p-f curve for a bearing contaminated by water, showing best-in-class potential performance for each condition monitoring data source. Specific vendors’ systems may perform worse than shown here.

Quick-and-dirty P-F curve for a bearing contaminated by water, showing best-in-class potential performance for each condition monitoring data source. Specific vendors’ systems may perform worse than shown here.

Example 2: Stator winding short circuit

A stator winding short circuit is a serious failure mode in electric motors, often caused by vibration, overheating, or electrical imbalance. Stator windings are insulated to prevent electrical shorts, but when the insulation degrades, shorts can occur, leading to severe motor damage.

Vibration sensors are effective at detecting the root cause of winding short circuits. For example, heavy vibrations can loosen the stator coils, which then rub against the motor’s metal housing, further wearing down the insulation. Early detection of excessive vibration allows for corrective action before the stator coils degrade.

Electrical sensors play a crucial role in monitoring stator winding health. These sensors can detect voltage imbalance or overvoltage, both of which contribute to insulation failure. Additionally, electrical sensors can monitor partial discharge, one of the earliest signs of insulation degradation. By catching these issues early, electrical monitoring prevents major short circuits, saving costly motor replacements.

Temperature sensors are useful for detecting overheating in stator windings, especially if the motor is operating above its rated load or has ventilation issues. However, as with bearing failure, temperature sensors typically indicate damage after it has occurred, making them more suited for tracking real-time conditions than providing early warnings.

Quick-and-dirty p-f curve for transient voltage unbalance, showing best-in-class potential performance for each condition monitoring data source.

Quick-and-dirty P-F curve for transient voltage unbalance, showing best-in-class potential performance for each condition monitoring data source. Specific vendors’ systems may perform worse than shown here.

Example 3: Cavitating pump

Cavitation is a common problem in pumps, where bubbles form in the liquid being pumped. When these bubbles collapse, they generate shock waves that can damage the pump impeller and other components over time. If left undetected, cavitation can lead to costly repairs and pump failure.

Vibration, electrical, and stress wave sensors are highly effective at detecting cavitation. These sensors can pick up the shock waves generated by collapsing bubbles, signaling that cavitation is occurring. For example, vibration sensors can detect the impact forces of cavitation, while stress wave sensors can capture the acoustic emissions caused by the bubble collapse.

Electrical data can also help by analyzing the pump’s real-time performance. For example, cavitation often occurs when a pump operates too far outside its ideal range, as indicated by its performance curve. Electrical sensors can track changes in pressure and flow, providing real-time information on when and where cavitation is likely to happen.

While temperature sensors are generally not effective for detecting cavitation, some studies suggest that infrared cameras can detect minimal temperature changes caused by cavitation. However, this requires a direct line of sight to the liquid, which is impractical in most industrial setups. Lubricant analysis is similarly limited, only detecting cavitation damage after it has affected lubricated components like seals and bearings.

Quick-and-dirty p-f curve for pump cavitation, showing best-in-class potential performance for each condition monitoring data source.

Quick-and-dirty P-F curve for pump cavitation, showing best-in-class potential performance for each condition monitoring data source. Specific vendors’ systems may perform worse than shown here.

Conclusion: a smarter approach to machine maintenance

Condition monitoring technologies offer a proactive approach to maintenance. By detecting faults early, you can schedule repairs, reduce downtime, and extend the life of your equipment. Whether you use electrical signals, lubricants, stress waves, temperature, or vibrations, the right technology can help ensure your machines stay running smoothly.

A quick recap of some major pros and cons of each data source of condition monitoring, for easy reference.

A quick recap of some major pros and cons of each data source of condition monitoring, for easy reference.

Book a demo

Get professional advice on how to monitor hard-to-reach assets from the motor control cabinet.

Talk to an expert

Monitor your hard-to-reach assets from the motor control cabinet

30 mins, no obligation

Experience our product firsthand

ABB_logo
Nobian-Logo
yorkshire-water-logo-white
thames-water-logo (1)
Schneider_Electric_logo