Don't miss MaximoWorld 2024, the premier conference on AI for asset management!

Experience the future of asset management with cutting-edge AI at MaximoWorld 2024.

Sign Up

Please use your business email address if applicable

Machinery Fault Diagnosis

This article is the result of three years’ experience of corrective maintenance (CM) experts in a petrochemical plant trying to explain implementation methods, obstacles, advantages and disadvantages of acoustic emission technique. It compares the vibration and acoustical data gathered from different machines in order to find similarities and differences.

Implementation Method

For implementation of AET, a variety of instruments have been introduced to industries. The measuring device in this petrochemical company is able to record the sounds and transfer the frequency range from 15 kHz to 197 kHz to the human hearing frequency range of 20 Hz to 20 kHz. This device is compatible with both sound and ultrasound sensors; the former evaluates ultrasound in terms of dBμV, while the latter evaluates sound pressure level in dB terms.

To utilize this device, measuring points of ultrasound should be marked on machines to minimize possible errors. Hence, data can be acquired, stored and analyzed. Subsequently, any abnormal evidence, such as an increase in measured data, can cause machinery to go under precise supervision. There are some notes related to data gathering that should be considered:

  1. Data acquisition should be done preferably by a specified individual due to applied pressure to the sensor.
  2. Measuring points and time interval should be carefully respected.
  3. The amplification setting in the instrument.
  4. The effect of noise in a noisy environment.

It’s highly recommended that acquired acoustic and vibration data be analyzed together. This may lead to finding some similarities, as well as differences, between AET and vibration. Likewise, before issuing any work order for dismantling a machine after tracking the fault, AET data should be gathered for future analysis.

To make acoustic data quantitative, it’s necessary to calculate some acoustical parameters. These parameters help verify the severity of the defects so necessary action can be taken to manage the lifetime of the defected part. Parameters are well used for detection of faults in roller element bearings and life management of them as well.

There are various parameters for evaluating an acoustic wave:

1. RMS: This is defined as the formula:

Where ΔT is integration time and N is the number of discrete data within ΔT.

2. Integral: Numerical integration of a certain part of the signal.
3. SNR: The ratio between noise and overall signal power. SNR is defined
as the power ratio between a signal and background noise, where A is
RMS of amplitude.

SNR is often expressed using logarithmic dB due to the wide dynamic range
of signals.

4. MARSE: Stands for measured area under the rectified signal envelope,

which can be clearly seen in Figure 1.

Figure 1: An illustration of MARSE

5. Kurtosis: A parameter that describes how a probability density function is distributed. Kurtosis can be defined using the formula:

Where x is signal, μ is average and σ is deviation.

6. Spectrum: One of the most popular analysis methods using Fourier transform function.

7. SVL: Stands for sound voltage level and is calculated using the formula:

Where V0, reference voltage, is considered 1μV.

Disadvantages of AET

There are some obstacles users face with acoustic emission technique.

They are:

  1. The effect of noise on recorded sound waves.
  2. The importance of how and in which direction a sensor should be held.
  3. Difficult analysis of recorded sound waves.
  4. Exaggeration of problems.
  5. Disproportionality between defect development and some AET parameters.

At the beginning of implementing AET in machine diagnosis, once the SVL was increased, it seemed that a serious defect existed in the equipment. Accordingly, several electric motors (about 3-4 motors) were dismantled in order to check and replace the bearings. Although the ultrasonic sound level was very high and there seemed to be a severe defect in the bearings, the only problem discovered was a false brinelling. There was no need to change the bearings (Figure 2) at that time and it was able to be in service more. False brinelling happens in these cases:

• The machine is stored in a warehouse for a long time, exposing it to the influences of humidity.

• A machine is out of service for a long time period, causing induced vibration

Figure 2: False brinelling effect on bearing

Advantages of AET

One of the most important advantages of AET is the early detection of faults in machines. Other advantages are:

  1. Insensitive to structural resonance;
  2. Detection of growing cracks;
  3. Low needed test time;
  4. Evaluating the lifetime of parts.

As mentioned before, early detection of faults, especially in roller element bearings, is one of the most important advantages of the acoustic emission technique. This will be discussed later in this article.

Study and Findings

Air cooled exchangers (ACEs) are the machines under supervision of AET in the Zagros Petrochemical Company (ZPC) plant. Figure 3 shows an installation schematic of these coolers.

Figure 3: Diagram of air cooled exchangers in ZPC plant

The ultrasonic sound level of these electric motors is acquired monthly. Accordingly, these motors are categorized into four groups based on the sound voltage level. They are:

Group 1: SVL<20 dBμV
Group 2: 20<SVL<30 dBμV
Group 3: 30<SVL<40 dBμV
Group 4: SVL>40 dBμV

Table 1 show the AET parameters values of 11 studied coolers. Note that the SVL is ascending downward.

In Figures 4 and 5, the values of various parameters versus the changes in SVL are shown. It's implied from this curve that as the SVL increases, the values of Kurt and SNR decreases. Likewise, increasing the values of SVL results in increasing the RMS and absolute values of integrals (Figure 5). Note that these parameters are calculated in a certain time period (10 seconds).

This phenomenon was seen in several machines and the same results obtained as well.

Bearing deterioration consists of several steps or phases:

Initial stage - increase in ultrasound level.
Second stage - large increase in ultrasound.
Third stage - very high increase in ultrasound.
Final stage - gradual decline in ultrasound.


Figure 4: Waveform parameters' changes trending                  

Figure 5: Waveform parameters' changes trending

In Figure 6, trending of ultrasound levels (SVL) in the bearing of an electric motor is shown in a bar chart in which an envelope curve has been set along with bars.

This process is also common in vibration analysis. Figure 7 illustrates the trend of bearing condition unit (BCU) in a bearing at the same time period.

Note that fault diagnosis should not be done using AET alone. In fact, defects have to be detected early using AET and controlled and monitored in order to postpone bearing replacement as much as possible.

Figure 6: Defect developing in a bearing versus SVL

Figure 7: Developing defect in a bearing

Statistical Investigations

ZPC includes two separate methanol plants. There are three sets of air cooled exchangers, with a total number of 106 fans in each plant (Figure 8). These fans are electric motor driven, in which power is transmitted by timing belts. Because of its vicinity to the Persian Gulf, ZPC is located in a hot and high humidity coastal area. As such, two main parameters play important roles in the operation of these machines: humidity and temperature (Min: 15°C and Max: 50°C).

Humidity: The relative humidity percentage in this region varies from 20 to about 95 percent during a year. Moreover, injected steam into a tower (T- 1501) (Figure 8), which comes out from the top, dispenses around the ACE and enters into electric motors due to an existing draft at the bottom of the air cooled exchanger.

Temperature: All the electric motors are erected vertically below the exchangers, so the drive end (DE) side of the bearing is too close to the finned tubes exposed to high temperatures.

According to the defined routine, ultrasonic sounds of electric motors are recorded, saved and analyzed monthly. The results of these data are analyzed statistically as well. Figure 8 shows the plant layout and potentially contaminant machines that are affecting the air cooled exchangers.

According to Figure 8, the cooling tower in plant 2 and the T-1501 in both plants are the main contaminant equipment. Likewise, the direction of wind in this region causes the outlet vapor from the T-1501 and the humidity from the cooling tower to move toward the ACEs and enter them.

Figure 8: Plant layout

These fans are equipped with two lubricating nozzles through which grease can be injected into the bearings. One of these nozzles is exposed to the polluting substances. These polluting substances can be entered, along with grease, into the bearing and cause deterioration of it.

Naturally, the electric motors closer to the source of the contaminants have the worse conditions. Hence, some information was gathered through the computerized maintenance management system (CMMS) about the electric motors, including:

  • The number of routines issued for regreasing (Routine A).
  • The number of routines issued for replacing the belt, leveling the skid, adjusting the belt tension and so on (Routine B).
  • The number of routines issued for replacing the bearings (Routine C).

SVLs for all electric motors are also recorded using an acoustic emission (AE) instrument.


The results of the collected data showing the overall conditions of the electric motors are shown in Figure 8 using different colors. The red color refers to the defected machine, the yellow color represents those electric motors whose bearings are in the initial stage of defect and the green color points out a normal condition. Analyzing the data proved these results:

  • The closer the electric motors to the contaminant equipment, the more vulnerability. As demonstrated, the red circles show defective fans, which are closer to the cooling tower.
  • The electric motor’s DE side bearing is more vulnerable than the non-drive end (NDE) side due to the exchanger’s high temperature, as well as pollution.
  • Since there is no cooling tower in plant 1, no regular deterioration was seen on the ACE- 3002 in this plant.
  • Although plant 1 was commissioned about three years earlier than plant 2, the number of defected electric motors in plant 2 are more than those in plant 1. In other words, the cooling tower is the main contaminant equipment in plant 2.


Acoustic emission technique is a valuable technique in machinery fault diagnosis, but unfortunately, is not used widely in industrial plants. Many written papers related to AET are the result of lab investigations. Although it cannot be claimed that without implementing AET machine diagnosis will face problems, but using this valuable technique will remarkably help with more accurate diagnosis and defect tracking. However, it should be noted that the nature of AET is somehow investigative not only in labs, but also in production plants.

Timely fault detection is one of the main advantages of acoustic emission technique. Sometimes, an equipment fault, especially in roller element bearings, can be detected, monitored and tracked several months before it will be seen in vibration spectrums. It could be said that there are some techniques, such as the envelope technique, which helps detect the fault in a timely manner. But the fact is that AET is famous due to its high sensitivity. Once a fault is timely detected by AET, some maintenance activities can be executed in order to manage the defect growth process and reduce its decline.

One of the main ways for reducing defects in anti-friction bearings is timely regreasing. Of course, the amount of injected grease into the bearings is of the essence. However, regreasing intervals also can be defined using acoustic emission technique.


The authors dedicate this article to Mr. Aliakbar Abbaspour, maintenance manager at Zagros Petrochemical Company, whose managerial method and strategy in maintenance is knowledge based and admirable.


1. Mindess, Sidney. Handbook on Nondestructive Testing of Concrete, Chapter 16, Acoustic Emission Methods. Boca Raton: CRC Press, 2003.
2. Hamstad, M.A., O'Gallagher, A., Gary, J. "Effects of Lateral Plate Dimensions on Acoustic Emission Signals from Dipole Sources." Journal of Acoustic Emission, Volume 19, 2001.
3. Hort, Filip, Mazal, Pavel. "Application of Acoustic Emission for Measuring of Contact Fatigue of Axial Bearing Engineering." Journal of Engineering Mechanics, Volume 18, 2011: 117-125.
4. Ríos-Soberanis, C.R. "Acoustic Emission Technique, an Overview as a Characterization Tool." Journal of Applied Research and Technology, Volume 9, 2011.
5. Journal of Applied Research and Technology, Volume 9, 2011: 367-379.
6. D. Mba. "Acoustic Emissions and Monitoring Bearing Health." Tribology Transactions, Volume 46, 2003: 447-451.
7. He, Yongyong, Zhang, Xinming, Friswell, Michael I. "Defect Diagnosis for Rolling Element Bearings Using Acoustic Emission." Journal of Vibration and Acoustics, Volume 131, 2009.
8. Kaphle, Manindra, Tan, Andy, Thambiratnam, David, and Chan, Tommy. "Acoustic Emission Technique - Opportunities, Challenges and Current Work at Queensland University of Technology." Smart and Intelligent Systems Proceedings, 2011.
9. Eshleman, Ronald L. "Machinery Vibration Analysis: Diagnostics, Condition Evaluation and Correction," Volume 2. Vibration Institute, 2002.

Hamid Karimi is a maintenance engineer for Zagros Petrochemical Company located in Pars Special Economic Energy Zone(PSEEZ). Mr. Karimi has worked for Shiraz Petrochemical Company and TEIF Consulting Engineering Company prior to joining Zagros in 2003. He has a BSc in Mechanical Engineering and a MSc in Construction Management.

Mohammad Moshtaghi is a mechanical engineer at Zargos Petrochemical Company located in Pars Special Energy Zone (PSEEZ). Mr. Moshtaghi has a variety of work experience in industrial projects, including cement construction and commissioning, as well as in petrochemical and terminal & tanks companies. His expertise is in the field of machinery maintenance and condition monitoring programs.

ChatGPT with
Find Your Answers Fast