Breaking Down Condition Monitoring

Jonathan Trout, Noria Corporation
Tags: condition monitoring

Condition Monitoring is the measuring of specific equipment parameters, noting signs of any significant changes that could be indicative of an impending failure.

Condition Monitoring  

What Is Condition Monitoring?

Condition monitoring is defined as the measuring of specific equipment parameters, such as vibrations in a machine, its temperature or the condition of its oil, taking note of any significant changes that could be indicative of an impending failure. Continuously monitoring the condition of equipment and taking note of any irregularities that would normally shorten an asset's lifespan allows maintenance or other preventive actions to be scheduled to address the issue(s) before they develop into more serious failures.

Condition monitoring is a big component of predictive maintenance. The data collected from condition monitoring over time provides valuable information about the current and historical state of an asset. This evolution of a machine can be used to anticipate how the asset will perform over time and how it might degrade, allowing for the scheduling of maintenance based on these predictions. This is known as predictive maintenance – maintenance based on what failures may occur and what maintenance should be scheduled to prevent such failures from occurring.

Condition monitoring techniques are typically used on rotating equipment (gearboxes, reciprocating machines, centrifugal machines, etc.), backup or secondary systems, and other machinery such as compressors, pumps, electric motors, presses and internal combustion engines.

There are two common methods used for condition monitoring:

  • Trend monitoring: Trend monitoring is the continuous, regular measurement and interpretation of data. It involves choosing a suitable and measurable indication of machine or component deterioration and studying this trend to figure out when deterioration goes over a critical limit. For example, trend monitoring is used for routinely tracking airplane engine data to detect and diagnose abnormalities in engine performance, hopefully preventing secondary, more costly damage.
  • Condition checking: Condition checking involves taking a periodic check measurement with an appropriate indicator while a machine is running. The information from this method is then used to measure the condition of the machine at a given time. An example of condition checking could be using an oil sight glass like a condition monitoring pod (CMP) to check the condition of a machine's lubricant in real time.

Condition monitoring via these two methods provides an inside look at how your machines and/or components are currently operating and, over time, offers a historical account of machine health.

Benefits of Condition Monitoring

Unsurprisingly, condition monitoring can lend itself to many benefits, including decreased maintenance costs, reduced downtime, extended asset life and cost savings on prematurely changed resources. For example, your condition monitoring system measures the amount of noise produced by a component. Over time, you notice a trend of machine failure soon after the noise level on the component reaches a certain level. Because you have a condition monitoring system in place, you can now set an alert on that component when it hits that noise level, which, in turn, lets maintenance personnel know they might want to consider replacing the component.

Modern technology has taken condition monitoring online (as will be discussed later), so internet-enabled and wireless-connected sensors and software provide real-time measurements around the clock. These measurements are relayed to integrated software for analysis and the storing of historical datasets.

Condition monitoring tools that spit out data in real time enable you to determine the root cause of an issue quicker, and wireless sensors on assets automatically connect employees with real-time data via remote access using smartphones or tablets.

Condition monitoring lets your facility go from a reactive approach to more of a predictive maintenance program. Once in place, condition monitoring provides you with 24/7 measurements, showcasing a clear picture of the health of your machines without adding additional labor.

Downsides of Condition Monitoring

Condition monitoring systems rely on visual data gathered from multiple sensors integrated with a software system. This means an added cost of purchasing and installing these sensors, as well as purchasing the tools necessary for condition monitoring (vibration analysis, infrared thermographers, etc.). There's also the added cost of training employees to use condition monitoring technology accurately and effectively.

Additionally, condition monitoring sensors might have trouble working properly under especially harsh operating conditions. Such conditions can damage sensors, forcing you to replace them on a more regular basis than anticipated.


Condition Monitoring Techniques

Condition monitoring techniques are standardized through ISO and the American Society for Testing and Materials (ASTM). ASTM outlines a variety of standards, mostly dealing with condition monitoring for in-service lubricants, while ISO standards 13372, 18436, 17359 and 13381 (among others) specify the guidelines for condition monitoring and diagnostics of machines.

Below are the most common techniques used for gathering data on the current condition of machinery.

    Condition monitoring vibration analysis
  • Vibration analysis. Vibration analysis is a process for measuring the vibration levels and frequencies of machinery and then using that information to analyze the machine's health. Vibration analysis can detect issues like imbalance, bearing failures, mechanical looseness, misalignment, resonance and natural frequencies, electrical motor faults, bent shaft and even cavitation. It's been estimated that vibration warnings can provide up to three months of lead time before an actual failure occurs. The link above provides an in-depth look into vibration analysis.
  • Oil analysis. Oil analysis is used to routinely analyze the health of machinery lubricants, oil contamination and machine wear. The oil analysis process includes moisture analysis, particle counting, elemental analysis, acid/base numbers, measuring viscosity and using Fourier transform infrared (FTIR) spectroscopy to determine several other parameters. For example, a spectrographic oil analysis test can break down the chemical composition of oil to predict possible issues. High levels of silicone and aluminum in oil reveal it is contaminated with dirt or grit (aluminum silicates), while high levels of iron indicate wear components. The link in this section offers an overview of oil analysis, including what it measures and what to look for in an oil analysis report.
  • Condition monitoring thermograpghy
  • Infrared thermography: Infrared thermography is the process of using a thermal imager to detect radiation (heat) emitting from an object, converting it to temperature and then displaying that temperature distribution in an image. It's commonly used to monitor electrical and mechanical conditions of motors, bearings (abnormal friction), refractory insulation and locating gas, liquids and sludge levels. The primary goal of infrared thermography is to ensure machinery is running normally by detecting abnormal heat patterns within a machine which could indicate a defect or inefficiency. The link above explains the types of infrared thermometers and cameras, how to use and assess infrared thermography, and more.
  • Ultrasound: Ultrasound testing is useful for low- and high-speed mechanical applications and high-pressure fluid situations. For example, a digital ultrasonic meter measures high-frequency signals emitted from bearings and displays those signals in real time in decibels per microvolt (dBuV). Over time, this metric creates a historical picture to predict increases in friction, rubbing and other bearing defects. The dBuV metric is also employed to predict proper lubrication intervals. Ultrasound testing is often utilized alongside vibration analysis.

    Ultrasound typically is applied in the shock pulse method (SPM) for condition monitoring – a technique using signals coming off rotating bearings as the baseline for efficient monitoring of machines. For example, imagine a metal ball hitting a metal bar. When the ball contacts the bar, a pressure wave spreads through both materials. The pressure wave is quickly damped out (transient). When the front of the pressure wave hits the shock pulse transducer, it causes a damped-out, back-and-forth movement of the transducer's mass. When the oil film on a bearing is thick, the shock pulse level is low (showing low peaks); when the level increases, the oil film thickness is reduced.

  • Acoustic emissions: ISO 22096 outlines the monitoring of energy through acoustic emissions. Acoustic emissions are an aspect of vibration analysis but are vibrations at much higher frequencies than those detected during traditional vibration analysis. With acoustic emission testing, you're looking for high-frequency signals caused by cracking or impact, as opposed to the repetitive synchronous movement generated by vibration. Acoustic emissions are separate from transducer position, machinery speed or the configuration of rolling elements.

Condition Monitoring Types

There are a variety of machine condition monitoring types and techniques, each serving a different role for collecting data. Below are the most common types.

  • Offline condition monitoring. Offline condition monitoring generally is used for less critical or semi-critical assets where periodic scanning is good enough to observe the current condition. Offline monitoring is most commonly utilized with vibration analysis when a periodic check of the pulse will suffice for a less critical machine, and with oil analysis when oil is sent to a laboratory for testing. With offline oil analysis, some organizations use sampling kits to test certain oils' viscosity and water levels onsite for basic results. Semi-automated, in-house testing equipment also is available to test for things like wear metals, nitration, oxidation and additive depletion.
  • Online condition monitoring. Online condition monitoring is the continuous measurement of an asset by wirelessly connecting machine-mounted sensors to integrated software to provide real-time warnings for things such as vibration analysis, acoustic emissions, ultrasound and infrared thermography. Online condition monitoring can be highly customizable thanks to a variety of sensors and monitoring systems. Factors like machine type, bearing type, machine speed, machine components and machine elements should be considered when choosing an online monitoring system.

    Once the appropriate sensors are mounted on the asset in the correct spot, they can be wirelessly connected to a remote condition monitoring system where they will display real-time data. Most systems can integrate multiple types of sensor data (vibration, thermography, acoustics, etc.), so at any given time, you can get a snapshot of your asset's current condition. Online condition monitoring also lets you set up real-time alerts for remote devices or email.

    You may be familiar with the term "portable machine diagnostics," where portable equipment is used to read data from mounted sensors. This is another way to describe a type of online condition monitoring.

  • Route-based condition monitoring: Route-based condition monitoring is a technique where a technician records data intermittently using a handheld device like an infrared imaging camera. Often, this method is employed to create a trend pattern and determine if more advanced analysis is needed.

Detecting Trouble Using Condition Monitoring

Consider this scenario: You take your car in for its regularly scheduled maintenance. Two weeks later, it breaks down due to a completely different issue. Just like cars, machines are vulnerable to these random, unpredictable failures. Certain types of maintenance, like reliability-centered maintenance and predictive maintenance, are based on the principle that failure isn't always linear and requires analysis of several asset aspects to detect possible failure indications. This is why condition monitoring is so useful: it lets you monitor multiple facets at once using the techniques discussed above.

Condition Monitoring, the P-F Curve and P-F Interval

The P-F curve is a graph showing an asset's health over time to determine the interval between potential failure (P) and functional failure (F). Potential failure is defined as the initial point at which an asset starts deteriorating or failing. For instance, a history of recorded bearing failures could tell you that the bearing typically fails after its temperature exceeds 70 degrees. Functional failure is the point at which an asset has reached the limit of its usefulness and is no longer operational. For example, you have around five days from when the bearing temperature surpasses 70 degrees to when it fails. The P-F curve is set on an x-axis to measure time and a y-axis to quantify the asset's condition. In this example, you should be inspecting the bearing every two to three days.

Condition Monitoring P-F Curve

Condition monitoring plays a significant role in detecting the P-F interval of the P-F curve. The P-F interval is the time between an asset's potential failure and its functional predicted failure. The idea is that your inspection interval should be smaller than the P-F curve to catch a failure before it occurs. Using condition monitoring lets you gauge an asset's condition, maximizing the P-F interval. Monitoring things like oil sampling and analysis, acoustic emission analysis, vibration analysis and infrared thermography are all condition monitoring-based techniques to give you an inside look at a machine's current condition.

The method and frequency of monitoring make a difference in the length of the P-F interval, according to Dale Blann, CEO of Marshall Institute. Blann says technology-based online condition monitoring provides the greatest P-F intervals and is less disruptive than other types of inspections like offline inspections where machines are generally shut down.


Condition Monitoring and the IIoT

The industrial internet of things (IIoT) is essentially a network of interrelated devices on mechanical and digital machines that give you the ability to transfer data over a large network without needing human-to-human or human-to-computer interaction. Modern condition monitoring systems use the IIoT to integrate numerous types of monitoring software into one system in real time, from anywhere in the world and across multiple devices.

IIoT-connected condition monitoring systems enable organizations to easily monitor several facets of each asset and identify performance, detect issues and even automatically schedule maintenance based on preset limits. Some of the biggest advantages of IIoT-connected condition monitoring include:

  • Cloud storage: The IIoT leverages cloud computing, allowing companies to store large amounts of data in the cloud as opposed to storing data onsite or in a data center. This is a big advantage due to the constant stream of data generated by machines connected to online condition monitoring systems. For example, research shows that one wind turbine takes 2,000 readings per minute, which equals nearly one terabyte of data each week.
  • Sophisticated analysis: IIoT-based condition monitoring systems use machine-learning algorithms to form conclusions about things like the health of your assets and ways to improve the accuracy of diagnosis.
  • Ability to use data from multiple machines: A significant amount of data is needed for machine-learning algorithms to have enough information to create a predictive model. For example, close to 100 instances of bent shafts might need to occur to train a predictive model to identify vibration levels that lead to bent shafts, which could take years. Gathering vibration data from multiple machines of the same type simultaneously lets technicians collect the same amount of data in much less time. Additionally, gathering data from many machines increases accuracy and improves the success of the predictive model over time.
  • Less need for human activity: IIoT-based systems allow remote monitoring of hundreds of industrial machines from any location and on multiple devices. This is a substantial benefit for industries like electric power and oil and gas, as it makes monitoring remote installations like pipelines, offshore drilling rigs and sea-based wind turbine installations easier. IIoT-based systems can automatically collect, aggregate and disperse real-time data to technicians anywhere in the world.

When building an IIoT-connected condition monitoring system, a few things should be considered before sensors and other equipment are purchased. It's important to take into account the type of equipment you will be monitoring, the data variables (what information you want to collect) and how you'll use the data.

How often do you plan to review the data? Generally, the more frequently you need to review data, the more bandwidth/data storage is required. You can also purchase a system that allows you to set predetermined times for when data is reviewed. For example, maybe you only want to check a certain asset at the beginning of a shift and review the data twice a day but still receive alerts when the data exceeds the preset limits.

Do you have an internet connection and power capabilities in the area near your equipment? If not, that is an extra cost you'll need to factor into the overall budget.

Is your equipment indoors or outdoors? Outdoor environments can limit the ability to get an internet connection wherever the equipment is located. Additionally, outdoor settings inflict harsher conditions on sensors and other condition monitoring equipment, so you may need to consider weatherproof or more durable sensors.