Condition Monitoring and Assessment

for Rotating Machinery

Sundralingam Muthanandan and Khairul Anwar B. M. Nor

Initially, when rotating machinery was used in industry, the maintenance practice
was run to failure basis. This means that an equipment was being operated until it
was unable to continue its intended function. This was the situation when correc-
tive maintenance was later being introduced and implemented. As the technology
progress and the rotating equipment becomes more critical to operation and business,
preventive maintenance was introduced. The preventive maintenance task is usually
recommended by the Original Equipment Manufacturer (OEM). This maintenance
requires a fixed time base or operational hour-based intervention so that some of the
critical components at a higher probability of failure can be inspected and replaced
if required.
As time progress, rotating machinery becomes more critical to operation and more
complex machine is being introduced into the rotating equipment. This made the
reliability of the equipment becomes very crucial, equipment reliability started to be
tracked, and unplanned downtime to machines became unacceptable. As such, con-
dition monitoring technologies were evolved to assess the equipment health while
in operation in order to understand and detect any incipient and hidden failures.
These technologies enable some failure modes detectable early, and consequently,
timely maintenance can be executed to prevent costly corrective maintenance. This
technology, which is used to detect and predict failures, is also known as predictive
maintenance. Thus, most of the oil and gas industry limits the corrective mainte-
nance to non-critical equipment where the cost of corrective maintenance is more
economical and justifiable as compared to preventive and predictive maintenance.

S. Muthanandan (B) · K. A. B. M. Nor

PETRONAS, Kuala Lumpur, Malaysia
e-mail: [email protected]

© The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2019 1
S. A. Sulaiman (ed.), Rotating Machineries, SpringerBriefs in Applied Sciences
and Technology, https://doi.org/10.1007/978-981-13-2357-7_1
2 S. Muthanandan and K. A. B. M. Nor

Fig. 1 Condition monitoring (CM) technologies for rotating machineries

Introduction

Condition monitoring is the use of special devices to measure specific aspects of

equipment deterioration in order to diagnose the cause and recommend remedial
action. Condition monitoring involves technical tasks in which operating parame-
ters are periodically or continuously measured and recorded. The data will be later
analyzed by comparing and displaying the data so that it could support any decision
making for operation and maintenance of the rotating machinery.
Most of the industry practitioners and experts realize that unplanned downtime is
very costly and time-consuming. Failure detrimental to equipment integrity could
lead to organizational reputation and opportunity for value leakages. Therefore,
detecting developing insipient failures or issues before they lead to catastrophic
failures is one of the most challenging and imminent tasks. This detection not only
provides insight into the operating condition but also provides direction to mainte-
nance activities so that one can prevent unplanned downtime. With the development
of various types of condition monitoring techniques including advance techniques,
the detection and diagnosis becomes easier, enabling appropriate prescriptive actions
to be taken in a timely manner.
Condition monitoring technologies for rotating machineries in plants, as shown
in Fig. 1, mainly include the followings:
• Visual inspection and observation
• Vibration analysis
• Lube oil analysis
• Thermography—temperature distribution and contour analysis
• Ultrasonic inspection
• Motor current signature analysis
• Performance monitoring
• Equipment online operating parameters analysis
Condition Monitoring and Assessment … 3

The condition monitoring process briefly comprises the following activities:

• Measurement, recording, and analysis of rotating equipment parameters to gauge
rotating equipment health status.
• The best reference data for condition monitoring program would be the baseline
data where the equipment is at its best mechanical condition. Subsequently, the
existing condition or health of rotating equipment is then compared with baseline
data, and analysis is performed to determine if there are any incipient failures or
problems which are showing signs or patterns of degradation.
The technique in condition monitoring requires systematic measuring and trend-
ing of equipment parameters to assess the present conditions of rotating equipment,
comparing against previous conditions and predicting the equipment remnant life.
By doing so, remedial actions are taken only on as-needed basis. Condition monitor-
ing identifies problems to equipment early, enabling failure preventive action to be
taken before it escalates to a critical problem. It also allows the maintenance team to
plan for spare parts, manpower, and optimized operational window to schedule the
maintenance activities. Some of the key advantages of condition monitoring are:
i. Detection of health of machine while in operation
ii. Maintenance is only performed when required
iii. Significantly reduce major breakdowns
iv. Increase equipment reliability
v. Elimination of unnecessary costly time-based preventive maintenance
vi. Non-stock parts can be ordered in advance
vii. Detection of maintenance-induced problems, especially when it involves major
overhaul.

Visual Inspection/Observation

Visual inspection is one of the most economic and basic means of condition moni-
toring. This is one of the earliest condition monitoring techniques employed by the
technicians using multi-sensory monitoring and adjustment, such as color coding,
labeling, and descriptive criteria based on appearance, sound, touch, or smell. This
technique normally involves daily walkabouts by the operators at the equipment
to observe for signs of deterioration. Some of the observations are critical such as
mechanical seal leaks for centrifugal pumps, tubing leaks, external corrosion, and
coating deterioration. In addition, the operator’s visual inspection covers lube oil dis-
coloration, lube oil foaming, mechanical seal buffer or barrier tank oil discoloration,
casing/metal surface hot spot, and abnormal sound. These are very crucial machine
information which cannot be substituted by any other online data.
The expected machine information from the operator daily walkabout checks
should include the following:
4 S. Muthanandan and K. A. B. M. Nor

• Smell (odor issues, overheating equipment)

• Noise (abnormal noises such as knocking sound from reciprocating compressors,
rough sound generated from the fin fan cooler bearings, hissing sound from pres-
surized hot air leak)
• Visual (exhaust emissions for internal combustion engine such as blue or white
smoke)
• Touch with suitable glove or personal protective equipment (PPE), for detection
of vibrations.
A structured round does not require all equipment to be checked in detail every
shift. It is assumed that the frequency of checks will be based on equipment criticality
to ensure an even spread of work that allows an operator to perform consistently a
quality structured round.
A sample of visual observation checklist for gas engine-driven centrifugal pump
is shown in Table 1. A good checklist should have the safe operating envelope which
specifies the operating limits. This provides alert to the operators in case the reading
is out of tolerance.
These data can be reviewed with the supervisor in charge and can be used for
trending of any changes which could give an early warning sign of degradation.

Vibration Analysis

Vibration measurement is one of the best methods available for detecting and moni-
toring mechanical problems and health condition of rotating equipment. It is normal
for Rotating Equipment (RE) to vibrate during operation. Rotating equipment that is
operating in the best condition will have minimal vibration and noise levels. When
RE vibration and noise increase, they imply incipient failures and mechanical fault,
or sometimes process-induced failure. The root causes of the RE mechanical fault can
be easily identified through its spectrum pattern, frequency dominant peak/amplitude
(severity levels), and absolute/relative phase angle (direction of vibration).
Vibration is the movement of the casing and/or rotating component around a refer-
ence point. Vibration can be caused by numerous sources from design flaws, installa-
tion or assembly errors, manufacturing defect and maintenance or process/operation
influences. These forces increase vibration levels due to unbalance, misalignment,
bearing defects, aerodynamic forces, gear problems, and many others. Vibration data
can be acquired and analyzed in the following forms:
i. Increase or significant reduction of overall vibration value over time.
ii. Transformation of the overall vibration signal/data into various sub-elements
such as spectrum analysis using fast Fourier transform (FFT) technique.
iii. Utilization of basic vibration signal (waveform collected from X-Y proximity
probes) to produce orbit plot, shaft center line plot, etc.
Condition Monitoring and Assessment … 5

Table 1 Sample of site visual inspection/observation checklist for gas engine-driven pump
Gas engine Low limit High limit Eng A Eng B
Lube oil pressure
Lube oil level
Lube oil top up quantity
Speed (RPM)
Exhaust temperature
Engine lube oil temperature
Fuel pressure
Engine jacket water temperature
Visual inspection Observations
Governor linkages condition
Spark plug and cable condition
Engine skid mounting bolts condition
Engine skid perimeter condition
Sign of oil leak
Lube oil color
Any sign of cooling water leak
Exhaust smoke discharge condition
Exhaust pipe insulation condition
Breather hose condition
Engine sound
Exhaust smoke condition
Engine body condition
Sign of oil leak at turbo charger
Pump Low limit High limit Pump A Pump B
NDE bearing oil level
DE bearing oil level
Top up oil (consumption)
Suction pressure
Discharge pressure
Pump discharge flow
Visual inspection Observations
Any sign of oil or product leak
Lube oil color
NDE mechanical seal condition
DE mechanical seal condition
6 S. Muthanandan and K. A. B. M. Nor

iv. Utilization of basic vibration signal (waveform and spectrum collected from X-
Y proximity probes with keyphasor/tachometer) to produce transient plots such
as Bode plot, waterfall plot, and others.
The primary purpose of vibration measurement is to collect, store, and detect the
deviation from their baseline and compare with the safe operating limits (SOL).

Techniques of Vibration Analysis

With the advancements of today, digital analyzers are used to acquire data and anal-
ysis. This has enabled industries to utilize handheld data collectors and spectrum
(FFT) analyzer, thus making data collection easier, more reliable, safer and provides
higher levels of accuracy. Vibration magnitude is also known as “amplitude.” This
can be measured in terms of displacement, velocity, or acceleration.

Displacement

This measures the change of distance or position of an object relative to a reference

point. On machinery, it is usually expressed in the unit of mils or micrometers (µm),
and it is usually measured in peak-to-peak form. Proximity probes are used to directly
measure relative displacement of the shaft. The displacement transducer (probe)
detects proximity of conducting materials. The proximitor produces an eddy current
in the coil, and the overall level of the voltage is determined by the cable and probe
circuit impedance. The impedance of the circuit is varied by the distance of the
conducting material.
The dual plane arrangement shown in Fig. 2 is necessary to determine the average
shaft radial position. The displacement arrangement can also be used for position
measurement such as axial thrust position and as a pulse for speed (keyphasor).

Velocity

Velocity is the rate of change of displacement. It measures the rate of the displace-
ment change. Typical units for velocity are inches per second (in/s) or millimeters per
second (mm/s) and zero-to-peak. It is measured with a velocity transducer. Velocity
transducers have a mass such as a permanent magnet which in turn is supported by
springs. The springs and magnet are surrounded by a damping fluid and an electri-
cal coil. In some applications, the velocity probe with integrated piezo element is
used. The vibration data which is initially measured in acceleration is converted into
velocity by mathematical integration.
Condition Monitoring and Assessment … 7

Fig. 2 Proximitor location at bearing housing

Acceleration

Acceleration is the rate of change of velocity. It measures the total force required to
move the vibrating element in the opposite direction. Typical units for acceleration are
inches per second per second (in/s2 ) or millimeters per second per second (mm/s2 ).
In some cases, it is measured as gravitational acceleration, g which is usually given
as 9.81 m/s2 . The accelerometer is a piezoelectric transducer with charge amplifier,
which generates an electrical signal that is proportional to acceleration.

Vibration Value

To enable correct trending of data, the way in which the vibration is measured has to
be consistent. There are three ways of measuring the value. The ISO convention is
8 S. Muthanandan and K. A. B. M. Nor

to measure in RMS (root mean squared); however, it can also be measured in “peak
to peak” (pk-pk) or “peak.”

Trending

The overall vibration trending either via continuous online or periodic offline data
will show indications of potential mechanical fault. The change seen in the trending
shows mechanical incipient or degraded failures. So, the overall value trending will
indicate the rate of deterioration so that operators could make necessary planning
for rectification. As a practical approach, a baseline reading for vibration is taken
and values are representative of the in situ condition when the machine is at its best
mechanical condition. As such, the baseline reading is taken after a successful com-
missioning of the equipment or after a major overhaul. Thereafter, individual baseline
values provide the reference against the monitored parameters during operation. This
enables more reliable condition assessment to detect any changes to the machine.
Besides comparing against the baseline reading, the vibration readings are also
compared against safe operating envelope, namely the warning or alert zone and the
unsafe zone. Operating at the alert zone is allowed with increased level of monitoring
in place. The risk of operation beyond the alert zone is mitigated via operators’ fre-
quent monitoring and increase in the level of vibration data acquisition and analysis.
The operation at this region is not meant for long period but only as an interim oper-
ation until the equipment can be safely shut down for remedial action. Operating at
an unsafe level is prohibited as this could lead to detrimental damages to equipment.
If the equipment reaches this point, it is recommended to bring the equipment for a
safe shutdown and execute maintenance. The safe operating envelope is determined
based on the vibration standard or OEM limits for safe operation.
During the factory acceptance testing (FAT), the vibration limit such as overall
and discrete peak shall be referenced to the user specification. Examples are User
Technical Standards, if specified by the user the reference can be made to API stan-
dards such as API 610 for Centrifugal Pump, API 618 for Reciprocating Compressor
and API 617 for Axial and Centrifugal Compressor and Expander–Compressor. The
International Standard Organization (ISO) standards such as ISO 7919 Series and
ISO 10816 Series also provide the vibration limits and safe operating envelopes which
are normally used as a reference during operation. A sample of trending parameters
and safe operating limit samples is shown in Fig. 3.
However, the overall vibration data and trending will not tell the causes of the
high vibration. So, in order to perform troubleshooting, the data shall be transformed
into other forms such as spectrum analysis in order to know the cause, and thereafter,
to enable appropriate remedial action to be taken.
Condition Monitoring and Assessment … 9

Fig. 3 Sample of trending on vibration and safe operating envelopes

Spectrum Analysis

Spectrum is a graph that shows vibration magnitude (amplitude) against frequency.

The graph is produced as a result of transforming a signal from the time domain to the
frequency domain. The composition of a time signal is transformed into a collection
of sine waves. The procedure of doing the transformation is most commonly done
with an FFT analyzer.
The vibration spectrum analysis enables analyses of vibration amplitudes at var-
ious component frequencies. By doing this, one can identify and track the vibra-
tion occurring at specific frequencies. This is because rotating equipment problems
develop vibration peaks at specific frequencies, and this information is useful to
diagnose the contributory causes. These components will be reference to equipment
operating speed such as 1×, 2× running speed, gear mesh or impeller blade-pass fre-
quencies and sidebands, rolling element bearing frequencies, and sub-synchronous
energy associated with journal bearing instability.
Having variable duty machines, it is often found that vibration parameters are
sensitive to other operating conditions such as operating speed and/or load. Hence,
an increase in vibration amplitude is not necessarily indicative of deteriorating con-
dition. As such, these changes should not be flagged as a problem. So, the analysis
should not be concluded by merely looking into the vibration peaks and trends,
but shall be associated with other operating parameters before concluding it as a
mechanical problem.
Another essential function of the spectrum analysis is to understand the severity
of the problem. For instance, high-frequency reading will normally indicate anti-
friction bearing problem. So, by making reference to inspected bearing condition
during maintenance and to the level of high-frequency reading, one can establish
10 S. Muthanandan and K. A. B. M. Nor

Table 2 Vibration spectrum and equipment problem

No. Problem Vibration frequency Causes
1. Electrical problems Synchronous AC line Common electrical
frequency problem such as broken
rotor bars, eccentric rotor,
unbalanced phases in
multi-phase systems, and
unequal air gap
2. Oil whirl or oil whip or 0.4 × −0.48 × operating Low stability at bearings,
worn journal bearings, Speed excessive bearing
coupling damages, and clearance, sub-harmonic
resonance resonance, improper oil
properties
3. Imbalance, lateral critical 1 × operating speed Operating speed is at
vibration, resonance, and critical speed, casing
shaft bow distortion, rotor imbalance,
coupling resonance and
shaft or rotor bow
4. Misalignment, mechanical 2 × operating speed Reciprocating forces, low
looseness, bad belts (2 × stability at bearings,
RPM) and resonance excessive bearing and seal
clearance, improper oil
properties or worn bearings
and seals
5. Blade-pass frequency, Many times RPM, Operating speed is at
caused by aerodynamic and normally harmonic-related critical speed, casing
hydraulic forces, and gear frequency (no. of blades × distortion, low flow in
problem operating speed) centrifugal pumps and
short distance between
centrifugal pump impeller
and volute, reciprocating
forces, mechanical
looseness or bad gears
7. Bearing problems, Very high frequency Anti-friction bearing
especially anti-friction damages, support
bearings, gear damages, resonance, gear damages
and shaft rubs and internal rubbings,
cavitation, recirculation
and flow turbulence and
improper lubrication to
journal bearings

the severity of the damage. So, the maintenance inspection findings’ reference to
spectrum analysis can be further developed to establish specific damage severity.
This would be helpful in the prediction of failure and planning for early intervention.
The common issues that can be detected by spectrum analysis are listed in Table 2.
Condition Monitoring and Assessment … 11

Lube Oil Analysis

Every rotating machine is unique in its operating environment because of its installed
location, operating condition, fuels, and products encountered. The most critical
aspect for accurate result of lube oil testing should be through the application of
correct oil sampling and handling procedure. A specific plan shall be developed
based on Reliability-Centered Maintenance (RCM) or Failure Mode Effect Analysis
(FMEA) for actual frequency of sampling and types of testing based on the probability
of failure and severity. Maintenance plans for lube oil include sampling of the oil,
preservation, flushing, and also change out. The following subsections provide the
general guidelines on the machinery lube oil and condition monitoring strategy. In
addition to the lube oil testing, the lube oil consumption shall also be monitored
especially for gas turbine and reciprocating engine.

Categories of Testing

Three main testing methods are:

i. Visual Inspection method, which is a critical preliminary indicator of equipment
condition. This is done by allowing oil samples with settling time and later to
perform visual inspection.
ii. On-site Testing, which uses the sample kits from lube oil testing laboratories.
Having this instrument allows simple site testing where some critical testing
such as water content, viscosity, and Total Acid Number (TAN) are tested and
results can be obtained immediately.
iii. Laboratory Testing, which gives more accurate and advanced testing on the oil
sample. This requires the oil samples to be transported to the laboratories, where
more sophisticated testing is applied to provide detail information besides the
visual and site testing.

Visual Inspection

Through visual inspection, a sample is taken from the lube system, and the sample
is allowed to settle undisturbed for appropriate settling time.
• If the lube oil is found cloudy with a single layer, it is an indication of foaming. It
can be caused by contamination or deficiency of anti-foaming agents in the lube
oil.
• If the lube oil is found cloudy with a clear divided layer, it is an indication of water
contamination. This is normally contributed by failures such as following:
12 S. Muthanandan and K. A. B. M. Nor

– Seal leakage (especially steam turbine which eventually enters the bearing cham-
ber and contaminate the oil)
– Oil cooler leakage
– Long-standby (e.g., firewater engine where the water leaks out from the pump
packing and entered the bearing chamber)
– Clogged oil/mist eliminator system. This is generally applicable to the gas tur-
bine lube oil system designed based on API 614. When the oil/mist eliminator
is clogged, the mist is not removed from the lube oil system and thus remains
in the oil.
• If the lube oil appears dark or blackish, it is an indication that the lube oil has been
oxidized. The severity of oxidation depends on how dark the oil color is. Bearing
operation at high temperature or occurrence of bearing fire could be one of the
contributors.

On-Site Testing

Site testing requires portable kits which can be used to test the oil sample at site.
Some of the testings are as follows;
• Water content
• Acid content or TAN
• Viscosity
• Solid contaminants (particle counts)
Findings from the site testing can be used as basis for further testing at the labo-
ratory before a critical decision is made.

Lube Oil Laboratory Testing

Laboratory testing should be applied for samples found deteriorated or contaminated

from on-site testing for detail analysis. This is to optimize the number of testing
samples and the cost associated with the laboratory testing. The various types of
analyses are listed in Table 3, in which different types of oil testing are provided. In
addition to the lube oil analysis, it is important to monitor the lube oil consumption
rate as part of routine operator’s visual observation and records.
Condition Monitoring and Assessment … 13

Table 3 Oil testing and purpose

No. Oil testing Purpose
1. Viscosity testing Crucial for retaining oil film thickness, too
thick slow oil flow ability, too thin
accelerate wear
2. Flash point It tells the tendency of an oil to form a
flammable mixture with air
3. Water content (Karl Fisher titration) Water contamination badly affects the
lubricant by acting as a catalyst for
oxidation and depleting antioxidants.
Water promotes corrosion and formation
of varnish/sludge, which then leads to
filter clogging
4. Total acid number Due to aging, acid by-products are
formed. Strong acids are corrosive, while
the weak acids attack bearing surfaces
5. Oxidation stability test (RPVOT and Measures the remaining antioxidants, thus
RULER) the remaining useful life of the oil. Low
values is contribute by the formation of
free radicals in oil and this can lead to
formation of varnish and sludge
6. Spectrography Identify wear components by chemical
composition. It shows the start of
abnormal wear and the components which
are wearing. It assists in determining the
root cause of wear/failure
7. Ferrography Identify metal particles, size, and origin.
Oil with high % of large particles is
abrasive to bearings and clogs
filters/strainers
Note The testing method shall be referred to ASTM. The limits for the test result shall be referred
to original equipment manufacturer (OEM)

Lube Oil Consumption Monitoring

The lube oil consumption rate tells the equipment health for some internal degrada-
tion. For example, high lube oil consumption for reciprocating engine may indicate
worn out piston ring resulting to the lube oil leaks into combustion chamber. It may
also mean a leaking gasket between the cylinder block and cylinder head. If this is
noticed, other tests like compression test can be performed at the cylinders to confirm
if there are any issues with piston rings or gaskets. This can be even further verified
with the crankcase blow-by flow for any abnormalities. The exhaust gas with white
smoke may indicate for burning oil, and this is an additional clue in the diagnostic.
High lube oil consumption on gas turbine may indicate leakage of oil in the lube
oil system either via external piping or internal. It is very dangerous if the cause of
high lube oil consumption is not known. Any form of internal leakage in gas turbine
14 S. Muthanandan and K. A. B. M. Nor

shall be paid serious attention because there are many hot components that could be
an ignition source for a potential fire occurrence. Thus, the operation teams are highly
recommended to keep the lube oil top rate in their logs and inspect the gas turbine
condition inside the enclosure. It is recommended to digitalize this information for
trending and analysis purposes.

Boroscope Inspection

The boroscope inspection is mainly applicable to gas turbines. It is a critical

inspection to be carried out during the annual maintenance. Gas turbine boroscope
inspection is an internal inspection performed by a trained personnel to assess the
condition of the gas turbine from the air inlet through to the exhaust, using an instru-
ment specifically designed to examine the gas path via the access ports positioned
along the engine. In general, the followings are the benefits of boroscope inspection:
• Inspection of internal passages without major disassembly,
• Alert on degraded conditions so that early intervention can be planned prior to
catastrophic failure, and
• Contribution to reduce overall maintenance costs.
In order to perform boroscope inspection effectively, the unit must be washed
prior to inspection because one cannot see damage (i.e., sign of crack, coating ero-
sion) with dirty blades. These inspections are one of the critical diagnostic methods
for maintaining the equipment. Both rigid and flexible fiberscopes are used in con-
junction with specially formed guide tubes to inspect the internal stationary, rotating
components and to detect early signs of failure.
The boroscope inspection could provide more insight into incipient failures, which
could not be detected via other condition monitoring technique. For instance, minor
foreign object damage to the blades, crack initiation and propagation, corrosion,
as well as thermal deterioration, mild cracking, or distortion mostly could only be
detected via boroscope inspection and will not be apparent in vibration reading or
trending. This is especially for aero-derivative engine whereby the vibration mea-
surements are made at the casing, and therefore, minor or moderate blade damages
will not be strong enough to be transmitted to casing for detection.

Ultrasonic Inspection

The ultrasonic inspection uses ultrasound translators or convertors which provide

the facility to hear high frequency noise from equipment via headphone, gauge the
intensity of the sound from a handheld unit and perform analytical spectrum analysis
of the sound through the use of software. It is widely used in mechanical rotating
equipment applications such as:
Condition Monitoring and Assessment … 15

• Rolling element bearings—there are condition the bearing may slip inside the
housing, producing internal sound.
• Airflow disturbance for fans
• Reciprocating compressors valve leak
• Reciprocating engine suction or discharge valve leak
• Control valve leakages

Thermography

Thermography is the measurement of temperatures using an infrared device. The

most common application is in electrical transformers, switchgears, or switchboards,
but there are also useful applications for rotating equipment. It detects surface tem-
perature difference and hot spots. The most common areas that are applied are:
1. Reciprocating valves. High-temperature indicates valve passing or failure.
2. Cooling fan belts slipping condition may create heat that can be detected by
thermography.
3. Bearing temperature problems can be detected based on heat generated.
4. Steam leaks at steam turbine can be determined due to the heat.
5. Gas turbine external surface temperature profile for any form of leaks such as
from the fuel piping, burner, or exhaust joints.

Motor Current Signature Analysis

Motor current signature analysis acquires motor current and voltage signals without
interrupting production and analyzing the derived signal to identify various faults.
Data can be acquired from the motor control panel, enabling easy testing of remote,
inaccessible, or hazardous area motors. Similar to vibration analysis, an FFT analyzer
is required for converting the signals from time to frequency domain to analyze the
spectrum. Some of the capabilities in terms of failure detections are:
i. Stator winding health
ii. Rotor health
iii. Air gap static and dynamic eccentricity
Apart from electrical condition, it also can be used to determine the mechanical
condition such as the followings:
i. Misalignment/unbalance
ii. Load issues
iii. System load and efficiency
iv. Bearing condition
16 S. Muthanandan and K. A. B. M. Nor

Therefore, it is very suitable where there is no possibility to apply other con-

dition monitoring techniques such as vibration, lube oil analysis, and thermograph
for motor-driven units. Examples of critical application for rotating equipment are
submersible pumps.

Performance Monitoring

This technique often related to efficiency of the equipment or comparison against the
baseline performance of the equipment. In rotating equipment, it often applied to cen-
trifugal pumps, centrifugal compressors, and gas turbines. However, with advanced
technology, it has also been applied to reciprocating compressors using pressure–vol-
ume (P-V ) diagram.
As for centrifugal pumps, operating parameters such as suction pressure, discharge
pressure are used to calculate the pump head and with the measured flow, it is
plotted against the manufacturer tested curve to determine performance degradation
to the equipment. This is one of the indicators to determine the need for major
maintenance or overhaul. The NFPA 25 has stipulated the need for annual fire water
pump performance testing. Performance degradation of more than 5% shall need
further investigation to determine the cause of degradation. The typical centrifugal
pump performance curve is shown in Fig. 4.
Centrifugal compressor performance is also determined based on a similar
method. The gas turbine performance is a bit more complex as it is given by the
manufacturer based on the heat rate versus ambient air temperature graph. The heat
rate (commonly in kJ/kW h) is the reciprocal of gas turbine efficiency and therefore
is calculated based on:
Energy Consumption of Gas Turbine
Heat Rate  (1)
Output Work

where the energy input or consumption (in kJ) is based on fuel consumption, and the
output work is in the unit of kW h.
In some cases, the OEM provides the tested graph with relationship between
specific fuel consumption against the exhaust gas horsepower during the factory
performance testing. In order to execute the performance evaluation at site, one
needs the torque meter and the fuel metering device to plot the parameters into the
performance graph for comparison. The other methods to monitor the gas turbine
performance are as follows:
• Plot the turbine outlet exhaust gas pressure (corrected) versus exhaust gas tem-
perature (corrected). Plot this graph after the overhaul as a baseline, compare the
operating parameters against the baseline reference, and note for any changes.
This graph is good indication for overall gas turbine performance monitoring. An
example of the graph is shown in Fig. 5.
Condition Monitoring and Assessment … 17

• Plot the compression ratio of the axial air compressor of gas turbine versus speed
(corrected) after the overhaul or commissioning. Use this as the baseline for future
operating data, and note for any significant changes. This is a good indication on
the performance of the axial air compressor and can be used as an indicator for
blade washing activity.

Fig. 4 Typical centrifugal pump performance curve

Note: The term corrected value for gas turbine means the value is converted into ISO
condition (The ISO standard condition is also specified in API 616 standard). The
simple way to convert to the corrected value is by dividing the relevant data with
[(Temperatureambient (°C) + 273.15) /288.15]

Fig. 5 Typical GT exhaust gas pressure (corrected) versus gas turbine exhaust gas temperature
(corrected)
18 S. Muthanandan and K. A. B. M. Nor

• Plot the baseline graph variable inlet guide vane (VIGV) position against the engine
speed (corrected). Use this as the baseline for future operating data, and note for
any significant changes. This is a good indication of any malfunction in the VIGV
system such as electronic and mechanical problems.

Start Reliability and Trip Data

Not every equipment startup will be successful. There are various reasons that a
system will not allow for successful start. Therefore, the startup reliability needs to
be monitored and tracked to understand the start behavior of the equipment. Startup
reliability is tracked by:
No. of Successful Start
Start Reliability (%)  (2)
No. of Start Attempt

Tracking this information will indicate the reliability of the start system, and
later, this information can be used for benchmarking for comparison and system
improvements.
In addition, it is imperative to monitor the number of trips that occur on the
machine. It is an industry-known fact that there are trips which are caused by instru-
mentation called spurious trips. This occurs purely due to faults with instrumentation
device and no other genuine mechanical abnormal condition. However, every trip
causes additional mechanical wear out to the equipment due to transient condition.
For instance, typically every trip may add additional 10 h into the life of gas turbine.
Thus, the information needs to be tracked and monitored.
Besides this, for equipment such as gas turbines, the operators are recommended
to log the time duration taken from start of button till the equipment reaches its
normal operating condition. Similarly, this is also applicable when the equipment
requires a normal stop. Any deviation in the time will indicate that some internal
failures such as bleed valve or variable inlet guide vane (VIGV) problems, gas turbine
compressor fouling, changes in gas turbine load, and others which are taking place
in the equipment. This is one of the simplest and most powerful data for condition
monitoring of the gas turbine.

Implementation of Condition Monitoring Program

in Industry

In industry, condition monitoring is implemented as tactical strategy to assess equip-

ment health condition. As equipment becomes more complex and critical in nature,
the industry cannot afford low-reliability equipment. Condition monitoring tech-
niques becomes the fundamental technology to transform the old practice on run to
Condition Monitoring and Assessment … 19

Thermography

Ultrasonic

Performance
Vibration

Lube Oil Analysis

Good Concern Good Good Concern

Fig. 6 Condition monitoring health status for a gas turbine

failure to predictive maintenance. Equipment insipient failures are detected much

earlier before the equipment needs to be stopped on mechanical failure. The early
identification avoids severe damages and costly repair but also helps for a proper
planning on tools, manpower, and spares needed to execute the maintenance.
Vibration and lube oil analysis are the most popular techniques used in oil and
gas industry for rotating machinery. The other techniques are becoming popular
and gradually being applied for better prediction on equipment health and failures.
These techniques are performed by specifically trained personnel responsible for
their respective condition monitoring skills.

Integrated Approach for Real Time Condition Monitoring

Since condition monitoring activities are performed by a specific trained skill group,
the reports on equipment health generated by the individual based on the techniques
that have been used. It is important that all the health information that has been
obtained using various techniques is being integrated with equipment health, and
failure prediction is performed in an integrated manner. This would be the most
holistic approach in using the condition monitoring techniques. Figure 6 shows an
example of condition monitoring health status.

Risk-Based Maintenance Approach for Rotating Machinery

As for rotating equipment, condition monitoring techniques become a critical input

for risk-based maintenance or inspection. Risk-based inspection or maintenance is
the maintenance activity driven by risk rather than purely engineering-based data such
as condition monitoring. It is basically used to prioritize maintenance by focusing
maintenance resources toward equipment that has highest risk, while equipment with
lower risk is given less priority. Traditionally, the engineering team is responsible
for equipment uptime, and they focus on reducing the probability of failure without
20 S. Muthanandan and K. A. B. M. Nor

considering the consequence of failures such as production loss and maintenance

cost.
A risk assessment is performed by combining the probability of failure against
consequence of the failure to determine the risk ranking. The risk ranking is then
plotted against time, and the equipment with the highest risk is given priority by pro-
viding more resources to execute the maintenance and failure prevention activities.
By doing so, the risk is mitigated to as low as reasonable and practically possible
and the resources are optimized to reduce cost and meet business needs.
The probability of failure is obtained from the followings:
i. Historical failure data analysis such as previous maintenance data which can be
used in Weibull Analysis to determine the probability of failure with time. This
would be the appropriate method to determine probability of failure.
ii. Previous maintenance findings as described below in this section.
iii. Inspection and testing activities such as boroscope inspection.
iv. Condition monitoring data on degradation and failure prediction.
v. Failure analysis on equipment types. This analysis focus on failures such as
related to vibration, degraded performance, seal failure, and their probability of
failure with respect to time.
vi. Remaining life analysis as it is important on some critical components such as
gas turbine blades which has creep or fatigue life.
It should be noted that Weibull analysis gives a good co-relation on failure prob-
ability with respect to time on replaceable component or repairable components that
can be restored as good as a new one.
The risk assessment is performed by plotting the probability of failure with con-
sequence based on risk assessment matrix as shown in Fig. 7. The probability of
failure for rotating equipment casing can be determined by using technology such as
ultrasonic thickness measurement on components of the rotating equipment such as
casing and nozzle. Using these techniques, one needs to inspect the casing of rotat-
ing equipment at every opportunity maintenance and record the wear pattern and
rate such as erosion and corrosion. The casing thickness is measured at reference
point to assess the degradation and prediction of failure. This is a critical aspect of
rotating machinery maintenance and inspection as, generally, the rotating personnel
gives priority for the rotating components such as rotor while the casings are often
neglected.
In addition, during preventive maintenance such as overhaul, one needs to record
the observation during equipment dismantling. Critical information that needs to be
gathered includes the journal bearing clearance, wearing clearance, bushing clearance
for centrifugal pump. These readings are again compared against the last as-installed
or as-built reading to determine the wear rate. This wear rate information for the
equipment can be used as a guide to predict the future condition of the machine
based on operating hours. This information shall be used again to determine the
next scheduled maintenance at an optimal interval rather than a fixed time-based
approach. This information also provides the probability of failure with respect to
time and therefore shall be used as an input in risk assessment.
Condition Monitoring and Assessment … 21

Fig. 7 Determination of risk through the probability of failure versus consequence (Example)

Rotating equipment switching policy is very crucial to provide information on

hidden failures. This activity can be considered as testing activity for standby equip-
ment to surface any problems for keeping this equipment on standby for long periods.
This information will provide data on probability of the system failure into the risk
assessment matrix.

Remote Monitoring and Diagnostics (RM&D)

Remote monitoring and diagnostics (RM&D) is originated by the original equipment

manufacturer (OEM) to take some responsibility over the equipment reliability by
monitoring and trending the parameters transmitted online from the equipment via
customer data infrastructure. A group of experts working remotely at the OEM
diagnostic center is employed to analyze the data and predict the equipment health
condition. This system is intended to provide early warning sign for any equipment
internal failures so that action can be taken at a very early stage. By analyzing the
data, the OEM shall provide the action that needs to be taken to the operators. OEMs
who have access to big data information from similar equipment around the world
could use the data and knowledge for prediction of failure in more accurate and
precise manner.
As technology progresses, automated technology is available as software packages
in the market. They can be purchased and used even without extending the equipment
data to OEM. This technology saves internal man-hours on analysis as the system
can automatically read the deviation, early warning signs and notify the operator on
the next course of action. The algorithm for the automated monitoring and analysis
22 S. Muthanandan and K. A. B. M. Nor

can be developed in-house if sufficient expertise and knowledge of the equipment

are available in an organization.
The prescription function of the technique can be based on a list of common failure
modes associated with the equipment. The prescriptive action shall be established
to rectify each failure mode. This will help the diagnosis and prescription to be
performed in an automated manner. It reduces the man-hour utilization drastically
on monitoring and analysis while the workforce can focus on other useful activities
such as the development of long-term solutions.

Summary

The elaborated condition monitoring techniques are keys for predictive maintenance
approach and understanding of the risk of equipment during operation. The advance-
ment of technology has made condition monitoring data digitized thus enabling inte-
grated analysis to be performed for a high degree of prediction. Condition monitoring
technologies made equipment to be inspected based on the health diagnosis, and with
risk-based maintenance, the maintenance is prioritized to provide reliability and busi-
ness value to the organization. As time progresses, there could be other advanced
technologies which will be introduced to expand the diagnosis of equipment during
operation.

Bibliography

1. B. Quesnel,Best Practice for Using Oil Analysis in Lubrication Management (Noria Publication,
Machinery Lubrication, 4/2017, 2017)
2. A.J. Smalley, D.A. Mauney, Risk based maintenance of turbomachinery, in Proceedings of
the 26th Turbomachinery Symposium, Texas A&M University, Turbomachinery Laboratories
(1997), pp. 177–187
3. B.M. Basaraba, J.A. Archer,IPT’s Rotating Equipment Training Manual—Machinery Reliability
& Condition Monitoring (IPT Publishing and Training, 1995)
4. API Standard 616,Gas Turbine for Petroleum, Chemical and Gas Industry Services, 5th edn.
(Jan 2011)
5. API Standard 610,ISO 13709: 2009, (Identical) Centrifugal Pumps for Petroleum, Petrochemical
and Natural Gas Industries, 7th edn. (Sept 2010)