Boilers, Headers, Steam lines, Turbines, Feedwater Heaters and
Condensers are the main components inspected in a non-nuclear
power plant. The reason for inspection depends on the component
and its effect on plant operation. Boiler tubes and feedwater
heater tubes are inspected to avoid forced outages. Inspection of
turbine components is done for safety and operational reasons.
Steam lines are inspected for safety reasons. The inspection of
condenser tubes is primarily done to assess condition for
replacement decisions. Appropriate selection and application of
NDT techniques are key to the inspection of non-nuclear power
plants.
1. Boilers
Boiler tubes are usually the number one cause of forced
outages in a thermal power plant. There are a total of twenty-two
failure mechanisms in boiler tube failures (Lamping 1985). These
mechanisms are directly responsible for failure of boiler tubes.
1.1 OD Erosion, corrosion and overheating.
Outer Diameter (OD) wall loss in a boiler is caused by
erosion, fireside corrosion and short term overheating. Outer
diameter erosion is measured by ultrasonic thickness measurement.
Wall thickness measurements are performed with commercially
available ultrasonic digital gauges or portable ultrasonic
pulser-receivers using a dual transducer. Calibration is
performed on a curved calibration plate to simulate actual boiler
tube geometry. In addition, the alignment of the dual transducer
is maintained the same on both the boiler tube and the
calibration block.
1.2 Hydrogen Damage, Caustic Corrosion, Chemical Attack
Inner diameter (ID) pitting in boiler tubes may be caused by
hydrogen damage, caustic corrosion, chemical attack, etc. Because
this type of pitting is usually isolated, a careful examination
of the boiler tube length is required. Digital gauges are
severely limited when measuring tubes with ID pitting. Ultrasonic
scattering from ID pits will produce an undefined back surface
reflection signal and impair thickness measurement. When
measuring the thickness of a tube with ID surface corrosion, an
instrument with a CRT screen display is recommended. The screen
presentation will identify the back wall reflection for reliable
thickness measurement.
Hydrogen damage is one of the mechanisms that produces ID
corrosion. This damage is produced in the water wall tubes from
imbalance in water chemistry (Partridge, 1963). Tube bends,
circumferential welds and tube lengths across the burners are
most susceptible locations for such damage. Hydrogen damage is of
serious concern because it not only results in ID wall loss but
also a zone of decarburized material under the corroded area.
Ultrasonic thickness scanning is the first step towards detection
of corrosion caused by hydrogen damage. Since ID corrosion can be
caused by other mechanisms, hydrogen damage should be verified by
NDT methods. Decarburization caused by hydrogen damage reduces
the ultrasonic velocity. Velocity measurement technique should
therefore be applied for verification of such damage(Birring,
1989 ).
1.3 Cracking - Corrosion Fatigue, Stress Corrosion and
thermal Fatigue.
Outer diameter (OD) cracking in a boiler tube can be produced
through thermal fatigue, corrosion fatigue, etc. Visual Testing
(VT), Magnetic Particle Testing (MT), Penetrant Testing (PT) and
Radiography Testing (RT) are commonly applied for detection of OD
cracking. The depth sizing of such cracks can be performed
visually from the crack length and width or by eddy current
testing. Special send-receive eddy current surface probes are
recommended for crack sizing.
Inner diameter cracks with axial orientation may be caused by
stress corrosion and corrosion fatigue mechanisms. Refracted
shear waves are used to detect these cracks. Inspection is
performed by placing the transducer on the tube's OD surface with
the beam directed towards the area being inspected. A refracted
angle that maximizes reflectivity from the crack should be
selected. Maximum reflectivity from the crack is produced when
the incident angle on the crack is 45 degrees. This incident
angle should then be used to calculate the transducer's wedge
refracted angle. The calculated refracted angle is always less
than the incident angle.
Dissimilar metal weld (DMW) cracking occurs in welds that join
the low alloy steels with the stainless steel. These welds are
present in high temperature sections of the boiler, including the
superheater and reheater sections. The DMW cracking occurs along
or near the fusion line between the low alloy steel and the weld.
In addition to the crack, there can be presence of an oxide notch
that is commonly found on the OD surface of the DMW. An oxide
notch is initiated because of differences in the creep strength
between a weld metal and the low alloy steel heat affected zone
(HAZ). The presence of oxide notch is not an indication of crack.
Ultrasonics and radiography are two methods to inspect DMWs. When
properly applied, these methods can resolve DMW cracking from the
oxide notch.
1.4 Creep - ID oxide Scale
ID oxide scale can be produced when tubes in the reheater and
superheater have experienced high temperatures for extended
periods of time. The formation of ID scale reduces heat transfer
and results in a further increase of tube metal temperature. The
increase in ID scale and the associated tube metal temperature
promotes creep in the tube metal. Formation of creep results in a
loss of strength at high temperature. The final outcome of
excessive scale is a thick lipped, long term overheat failure.
Scale thickness measurements should be taken just upstream of
material upgrade and thickness upgrade locations. A history of
prior long term overheat failures should also be used to select
tubes for oxide scale inspection. The ultrasonic method for
measuring scale thickness is based on transmitting a wave through
the tube thickness. The thickness is calculated by measuring the
time difference between the signals reflected from the
steel/scale interface and the tube ID surface. Because of the
extremely small time difference, the application requires the use
of high frequency transducers in the 15 to 30 MHz range.
2. Headers
Headers are inspected for cracking in the welds and ligaments.
Weld cracking is inspected by using ultrasonics and Wet
Fluoroscent Magnetic Particle Testing (WFMT). WFMT is used for OD
cracking while UT is used for ID or midwall cracks.
Leaks in Headers can be caused by ligament cracking. Ligament
cracking is produced in the bore holes and the stub tube ID. The
cracking occurs due to cyclic events such as startup and
shutdowns, transients and thermal shocks. The hottest areas are
the most susceptible to ligament cracking, however, there can be
exceptions.
The most reliable method for detection of ligament cracking is
to first remove the stub tube and then perform penetrant
inspection. ID grinding is first done to remove the oxide scale.
This is followed by the wet fluoroscent penetrant method. The
inspection determines the length of the crack in the bore hole
region. The crack length information is the used to make decision
on the disposition of the header.
3. Steam Lines
Inspection of steam lines is done to detect cracking in the
welds. Bending loads can produce OD cracking in the
circumferential welds. WFMT is the recommended approach for such
an inspection. The inspection should also be done on the hanger
welds to determine their integrity.
High temperature creep can cause midwall cracking or ID
connected cracking in seam welds. Over a long period of time,
creep voids can grow to microcracks interlink and cause failure
of a long seam weld (Viswanathan, 1989). The inspection of long
seam welds gained significant coverage in technical literature
after their failures caused loss of life. The failure of a steam
line can either be a "leak before failure" or a
rupture. The type of failure depends on the length of the crack.
Cracks longer than the critical length can result in rupture.
Ultrasonic testing is the recommended approach for such
inspection. The inspection is done with refracted shear waves.
Because of the pipe curvature, proper selection of refracted
angles is key to this inspection. The refracted angle is always
higher than the incident angle at the crack.
4. Turbines
There are several components that are inspected in a turbine.
These include bore, disk keyway, disk blade attachment area,
blades, nozzles, casing, bolts, etc.
4.1 Bore
The mechanism of crack growth in a rotor bore is due to the
combined action of creep and fatigue (Viswanathan, 1989). Creep
is more prevalent in HP rotors that operate at temperatures
around 1000°F. Fatigue is more prevalent in LP rotors that
operate at lower temperatures. Sensitivity of the bore
examination depends on locations that experience the highest
level of stress and temperature. The hoop stress is higher under
the disks because of mass loading. The temperature is highest
under the control stage. Sensitivity of examination should
therefore be highest in the HP rotors at the ID bore surface
under the HP disks.

Three methods are commonly used for bore inspection: magnetic
particle testing, eddy current and ultrasonics. The first two
methods are limited to surface cracking. Magnetic particle is
performed by applying a circumferential magnetic field at the
bore ID. The circumferential field detects axial cracks on the
bore surface. The second approach for surface crack detection is
the eddy current method. Ultrasonic inspection is the only method
that can perform a complete volumetric examination. A combination
of transducer angles is used to perform the inspection. The
transducers are installed on a scanner and the data is recorded
on an ultrasonic imaging system. The detection sensitivity is
controlled by adjusting the scan step interval.
4.2 Solid Rotor
The main advantage of a boreless rotor is its lower level of
stress compared to the bored rotor. The lower level of stress
makes the boreless rotor tolerant to larger flaws. Because of
this reason, inspection procedures for boreless rotors are less
stringent. The inspection of a solid rotor is performed by using
a combination of L-wave transducers and S-wave transducers. The
L-wave transducers can detect flaws directly below the
transducer. However, rotor sections directly below the disks
cannot be inspected with the 0° transducers. These locations are
inspected using refracted S-wave transducers. A range of
refracted angles, between 40° to 70°, is used to assure a
complete volumetric examination. Inspection of boreless rotors
requires that the selected angles be able to inspect the entire
material volume of interest.
Transverse cracking in LP rotors initiates from corrosion pits
can grow during service by corrosion-fatigue. Transverse cracking
is easily detected by application of magnetic particle testing
(MT) on the rotor OD surface. The depth of the crack can be
measured using the ultrasonic tip diffraction method.
4.3 Disk keyway Cracking
The primary cause of disk keyway cracking is stress corrosion.
High stress concentration in the keyway region promotes growth of
this cracking. Because of the SCC mechanism, keyway cracking is
mostly observed past the Wilson Line. In some cases stress
corrosion cracking may be found before the Wilson Line if
condensation occurred during standby. Inspection of keyway
cracking is performed using a range of ultrasonic refracted
angles. A combination of transducer angles for each disk is
selected so that the entire length of the keyway can be
inspected.
Both pulse-echo and pitch-catch modes are used during
inspection. The pulse-echo mode is preferred as it is easier to
apply and interpret. Normally, the ends of the disks are
inspected in this mode. The middle section of the disk cannot be
inspected with the pulse echo mode. This area of the keyway is
inspected in the pitch-catch mode. In this mode, a transducer is
placed on each side of the turbine disk; one transducer
transmitting and the other receiving ultrasound. Alignment of the
transducers, in the pitch-catch mode, is very critical to assure
a reliable inspection.
4.4 Disk - Blade Attachment Area
The mechanism of crack initiation and growth in turbine disk
blade attachment (steeples) depends on three variables: the
operating temperature, stresses and environment. Creep is the
primary mechanism in HP and IP rotors. Stress corrosion
cracking(SCC), combined with fatigue, is the primary mechanism
for LP rotors. Initially, cracking in an LP rotor grows slowly by
SCC. When the stress intensity KI exceeds Kth,
crack growth is predominantly due to fatigue. Crack growth rates
in this mode are significantly high because of vibratory loads.
Generally, failure can be imminent when the threshold for fatigue
crack growth Kth is reached. Therefore it is important
that NDE inspections detect cracks before their stress intensity
reaches Kth.
The inspection methods applied to detect steeple cracking
depends on the geometry. Dovetail design (GE turbines) can be
inspected only by ultrasonics (Bentzel, 1993). This design does
not allow access on the surface for eddy current or magnetic
particle testing. On the contrary, side entry steeples
(Westinghouse turbines) allow access to the side surface. In
addition to ultrasonic testing, these disks can be inspected by
eddy current testing and magnetic particle testing. However,
ultrasonics is the only method that is capable of inspecting the
entire length of the side entry steeple under the blade.
Once the blades are removed, WFMT is the preferred method for
inspecting steeples. WFMT is performed with a yoke on each
steeple individually. The process is slow, but results in a
highly sensitive inspection. It can also get evidence of cracking
at early stages.
4.5 Blades
The failure mechanism of turbine blades is dependent on their
temperature, environment and stress state. Corrosion fatigue is
the major failure mechanism of blades in the next-to-last stage
of the low-pressure turbine. Creep blade failures are limited to
HP turbines. Cracking of blades occurs at the following three
locations: blade attachments, airfoil and tenon. The inspection
methods chosen for each of these locations depend on whether the
inspection is performed with the blade removed from the disk or
not.
Eddy current and magnetic particle are the two methods used to
inspect the blade attachment areas of side entry blades. Eddy
Current inspection is an attractive method since it can be
performed without removal of turbine from the casing. During
examination, port holes in the turbine casing are used to gain
access to the blades. An eddy current probe along with a fiber
optic probe are held on the end of a rod which is inserted in the
casing through a port-hole for inspection.
Once the turbine has been removed from the casing, the blade
attachments become directly accessible for inspection. Either
eddy current or magnetic particle testing may be used at this
stage. Magnetic particle testing is the preferred method because
it is faster. WFMT is performed with AC coils or a yoke.
WFMT is the most commonly used method for inspecting blade
length. The inspection is performed by magnetizing the blades
with AC coils. The AC magnetization allows a highly sensitive
examination on the surface, while leaving minimal residual
magnetism in the blades. Blade tenons are located at the tip of
the blades and hold the shroud. Cracking and failure of the
tenons may release the shroud and cause mechanical damage to
other blades. The only method available to inspect blade tenons
is ultrasonics. An ultrasonic transducer is placed on the tenon.
A flat surface on the tenon is required so that a contact with a
transducer can be achieved. Tenons without a flat face can not be
inspected unless they are ground flat.
4.6 Bolts
Creep-rupture and brittle fracture are two primary reasons for
bolt failures. The low toughness that leads to brittle fracture
is due to the inherent high strength of bolts. The failures are
usually initiation controlled. Hence, the failure time is very
short after the crack initiation.
Ultrasonic testing is the only method that can inspect bolts
without removing them from the casing. Two different ultrasonic
approaches are used for this inspection. A zero degree
examination is performed when the top surface of the bolt is
flat, since it allows placement of a normal beam transducer. But
when the top face of the bolt is not flat, an angle beam exam is
performed through the heater holes.
Cracking in bolts occurs only in the threads next to the
joint. These threads experience the highest level of stress. The
stress on the last thread at the end of the bolt is almost zero.
Therefore, the inspector should carefully investigate threads
right next to the joint.
4.7 Retaining Rings
The susceptibility of 18 Mn 5 Cr steels to SCC produces
cracking in retaining rings. Initiation of cracks in the
retaining rings occurs when moisture enters and settles on the
inner diameter (ID) surface. The initiation time of the cracks is
quite long. Nevertheless, once the crack has initiated, crack
growth can be quite rapid. The high crack growth rates limit the
application of NDE for crack detection. No effort is made to size
the cracks once they are detected. Repair or replacement actions
are initiated once a crack is positively detected.
Four methods are generally used when inspecting the retaining
rings: 1) visual testing, 2) fluorescent penetrant testing, 3)
eddy current, and 4) ultrasonic testing. Ultrasonic testing (UT)
is the only method that may be applied without removal of the
retaining ring, however, its detection sensitivity is limited. A
combination of adverse factors, such as high ultrasonic
attenuation and spurious geometrical reflection, result in the
low UT detection sensitivity. Visual, eddy current and
fluorescent methods can be applied after ring removal. Visual
inspection is accomplished using borescopes that detect moisture
in accessible areas of the ID surface. Evidence of moisture is an
indication of possible crack initiation. The penetrant
examinations are performed using the fluorescent penetrant
method. High sensitivity Lipophilic emulsifiers are used for
these inspections.
5. Condensers
The main reason for inspection of condenser tubes is to
determine the condition of the tubes. The information from the
inspection can then be used to make decisions on replacement of
all the tubes. The inspection of condensers is normally limited
to a 5 to 10 percent random sample of tubes.
The most common tube materials in a feedwater heater are
copper-nickel alloys, brass, titanium, stainless steel and
ferritic stainless steel. Pitting is the most common form of
damage in condenser tubing. OD erosion/corrosion is very common
in brass tubing. Tubes in the top row are susceptible to OD
erosion.
The inspection techniques for condenser tubes depend on the
material. Conventional eddy current is applied for
non-ferromagnetic materials such as:
copper-nickel alloys, brass, titanium and stainless steel
tubing. Conventional eddy current can, however, not be used on
ferromagnetic materials such as thin
ferritic stainless steel tubing. For such materials full
saturation eddy current technique is applied. Because of the long
length of the tubing, inspection of condenser tubing is done at
high speed pusher pullers.
6. Feedwater Heaters
Tube failures in feedwater heaters are one of the major causes
of forced outages in a fossil power plant. Inspection of HP
feedwater Heaters produces one of the highest cost benefits of
any NDT inspection in a power plant.
The most common tube materials in feedwater heaters are carbon
steel, stainless steel, brass and copper-nickel alloys. OD
erosion is the most common type of damage in carbon steel tubes.
The locations most susceptible to OD wear are the drain cooler
section and the desuperheating zone. Pitting can occur in tubes
made out of stainless steel and copper alloys.
The tubing in the feedwater heaters should be periodically
inspected to determine its condition. Depending on the rate of
degradation, an inspection interval of 3 to 6 years is
recommended. Selection of tubes for inspection is key to an
effective feedwater heater inspection. For a regular inspection
the plan should include tubes in the drain cooler section, tubes
in the desuperheating zone, tubes around previously plugged tubes
and some tubes at random. In addition to regular inspections,
inspection after a tube failure is highly recommended. A tube
failure is an indication of damage and impeding tube failures.
During such an emergency inspection, tubes around the leaking
tubes should be tested. Tubes with damage above certain level
should be immediately plugged. Such an approach results in
effective plugging and avoids future forced outages.
Conventional eddy current is applied for non-ferromagnetic
materials. Remote Field Eddy current is quite effective for
inspection of carbon steel tubing. Unlike conventional eddy
current, this technique is only sensitive to wall loss and not
pitting. However, pitting is not a problem in carbon steel
tubing. In addition to remote field, ultrasonic IRIS technique
can also be applied for inspection while the IRIS technique is
more accurate, it is slow compared to remote field eddy current
technique. In general Remote field is used for the normal
inspection and IRIS can be used for verification
7. Conclusions
There are a variety of components in non-nuclear power plants.
For each of these components, there can be different types of
flaws and damage. This may include cracks, pitting, material
degradation, etc. Because of this combination of component types
and defect types, several types of NDT methods have to be
implemented. A careful selection of NDT methods is necessary for
effective NDT of non-nuclear power plants.
Theo nde.com
|