Common Problems in Feedwater Heaters: Presented To FSRUG January 19, 2017

Common Problems in
Feedwater Heaters
Presented to FSRUG
January 19TH, 2017
Presented and Authored By:
Steven Schaefer
and
Paul Koss
1 | January 23, 2017 | © 2014 Curtiss-Wright
PROPRIETARY and CONFIDENTIAL

INTRODUCTION
What are the common problems in Feedwater Heaters?
• Vibration damage
• Wear at the Tube Support Plates
• Tube-to-Tube Wear mid-span
• Fatigue failures at either
• Steam erosion of the tubes, tube support plates and/or the shell
• Steam erosion / by-pass at the drain cooler end plate
• Liquid level control –
• Too low and steam is allowed into the lower pressure drain cooler and two phase flow can result which
can cause severe damage
• Too high and not only is there thermal performance lost but the liquid can flash to steam releasing high
energy or deflect the steam into the tubes causing vibrational failures
2 | January 23, 2017 | © Curtiss-Wright

INTRODUCTION
What are the common problems in Feedwater Heaters?
• Broken, failed or inadequate steam impingement plates
• Broken, eroded or failed TSP tie rods
• Broken, eroded or cracked partition plates
• Foreign material or loose parts in the shell side or tube side
• Stress Corrosion Cracking in the roll transition / tubesheet crevice in high
temperature / high pressure FWH’s
• Failures due to Denting from operation, transport and/or manufacturing of the
FWH
• Leaking or loose tube plugs
• Vent channel erosion can lead to tube failures

INTRODUCTION
What can be done to reduce the chance of FWH leaks?
• A risk based ET inspection schedule that will identify the problem before it becomes a BIG problem
• Review of the ‘as built’ design drawings to identify potential problem areas to be sure they are
inspected if the current program is not a 100% inspection plan
• Modify the ET inspection technique to be sure the potential problems can be identified
• If erosion or vibration damage is found in the drain cooler talk to Operations and the FWH system
engineer to review level control variations and verify the level control is functioning properly
• If failures are occurring at the DC End Plate look for erosion of the hole and possibly expand and/or
sleeve the tubes to get better sealing and change the local vibration characteristics
• If erosion is identified at the TSP’s then be sure to UT the shell and conversely, if UT identifies shell
erosion be sure to have the vendor look close at the TSP’s
• If denting is suddenly occurring or growing at the TSP’s perform an axial measurement of the locations
of the TSP’s to see if they are moving and/or locking the tubes

INTRODUCTION
What can be done to reduce the chance of FWH leaks? (cont’d)
• Perform ‘as found’ inspections when the channel head and partition plate(s) are removed for Foreign Material or
loose parts on the tube side and be sure your vendor is looking for and reporting them on the outside of the tubes
• If there is suddenly erosion, vibration, denting or what appears to be a loose part where the impingement plate is
expected to be then perform a video inspection of the area to be sure the plate has not broken loose
• Check the venting to be sure it is functioning properly
• It is reported that some utilities puncture the tube prior to plugging it to remove the possibility of pressure build up
and ejecting the plugs and/or stabilizing cable into the channel head
• Stabilizing cables can be used to prevent tubes that sever from flailing
• If FWH replacement becomes the only option perform an in-depth analysis of exactly what caused the failure and be
sure to compensate for that in the new FWH design but be careful not to create another problem with the change.
Consult with FWH design experts to discuss the possibilities.
• EPRI is currently working on a “HX Design Specification Guide” which may add some insight of what to ask for and
expect to receive

Feedwater Heater Design
 Feedwater heater design has been relatively unchanged for many
years. There are three basic types:
– Single zone – condensing zone only
– Two zone – condensing and drain cooling zones common to lower pressure
lower temperature nuclear steam cycles
– Three zone – de-super heating, condensing and drain cooling zones –
mostly found in higher pressure and higher temperature fossil steam cycles
 The main differences in FWH’s are due to overall costs and the
lower the cost the more chances of future design/material problems

In the old days
the S/G and
FWH’s were
combined.

This is a three zone FWH. The steam first enters the de-superheating zone, then passes
to the condensing zone and changes liquid and then out through the drain cooler zone.

It is not clear if this is a two zone or a single zone but this illustrates
the condensing zone of a large FWH with 3 extraction steam inlets.

This is what is called the skeleton. The tube supports and tie rods
are visible and the tubes should be inserted shortly.

New U-tubes in Mill After Electric Annealing

This would be a very small FWH but the segmented tube supports/baffles are
clearly visible while the tube insertion is in progress. The segmentation is for
cross flow for more efficiency. Note the ½ holes only make 180 degree contact in
the overlap area.

Vibration Damage
There are two basic types of vibration damage in FWH’s:
1) Vibrational wear at Tube Support Plates

Vibration Damage
There are two basic types of vibration damage in FWH’s:
2) Tube-to-Tube vibrational wear at mid-span

Vibration Damage
Area of tube
leaks and
inspection
results in lower
left corner. The
damage was
located in the
condensing
zone below the
extraction
steam inlet.

Vibration Damage
The ‘dry’ tube
ends were the
nine leakers.
The plant had
been at low
power so the
tubes that were
open to steam
side dried and
became very
easy to identify.

Vibration Damage
The suspected
leakers (dry tubes)
were inspected with
a video probe to
identify the nature of
the damage and it’s
exact location.
They were severed

at approximately
15’3” from the
tubesheet in the
condensing zone.

Vibration Damage
In this photo
vibrational wear at a
TSP is visible.
Due to the limited ET
inspection scope it
could not be
determined if Tube-
to-Tube wear was
also present but
given the complete
failures it was likely
and collateral
damage was
occurring.

Vibration Damage
This was part of the ‘as built’ design drawing and notice the extraction steam inlet located
near the middle of the heater and at 45 degrees to the left of center which is also the side
where the leakers were located. The steam would enter and travel down the left side and
deflect off and/or flash the liquid level at approximately the bottom row of tubes and up into
the tubes in the area of the leakers causing severe vibration damage.

Vibration Damage
A Cad program was used to illustrate the damage areas in this FWH and help identify
what the mechanism was.

Typical Damage Mechanisms – Tube-to-to-Tube Wear




Erosion of Internals
Steam flow can
erode away the
carbon steel tube
support plates
while leaving the
stainless steel
tubes undamaged.
However they are
then unsupported
which can lead to
failures from fatigue
and/or vibrational
damage.

Area of TSP
erosion.
Note the
large gaps
between the
tubes and
the TSP
holes.

Area of TSP
erosion. Note
the large
gaps
between the
tubes and
the TSP
holes.

Actual eroded
DCEP holes.
Note the large
gaps between
the tubes and
the holes.

Drain Cooler
End Plate,
Shroud,
Inlet Snorkel

So maintaining
level around the
bottom row of
tubes keeps the
snorkel
submerged and
reduces the
chances of two
phase flow.

View of a
thick (~3”)
Drain Cooler
End Plate

Showing
back face of
tubesheet of
a higher
pressure
FWH with a
small Drain
Cooler.

View of
Tubesheet
showing a
separate
Drain Cooler
Section.

View from U-
Bend end
with tubes
installed
showing the
separate DC
and the end
of the full
length
Shroud.

Prebuilt
Drain Cooler
Shroud
showing the
Inlet Snorkel.

View
showing
vent
channel
and vent
line.

View of vent
channel.
That is why
there are
‘blank’ areas
in the
tubesheet to
allow for the
channel to run
through the
length of the
FWH.

View of full
tube support
plate for the
condensing
zone, tie rods
with
protective
coverings
and vent
channel

A FWH that
has been in
service and
the venting
line has been
eroded
completely
so NO
effective
venting.

Extraction Steam
Impingement
Plate
Eroded Tie Rod
Broken Welds

ECT Analysis response to a nominal TSP

ECT Analysis response to an eroded TSP.
Note the smaller TSP signal and the deposit signal indicating the build up of CS deposit left from
the erosion.

Horizontal U-Tubed single

zone FWH with extraction
steam inlets at 12 o’clock.
The red tubes indicate
tubes recording possible
TSP degradation while the
blue tubes were inspected
with no damage of any type
recorded.

Cad drawing showing the locations that recorded erosion of TSPs

Tube-Pro 3D Modeling
Tubes with TSP Degradation

Identified on Tubesheet

Locations at Tube Supports that

indicated TSP Degradation

Close up of area showing TSP Degradation

Liquid Level and Steam By-pass into Drain Cooler
The drain cooler is lower pressure than the condensing zone so the liquid level
must be kept above the drain cooler inlet snorkel. The normal level is just below
the bottom row of tubes. If the level drops then steam is allowed into the drain
cooler it will flash liquid to steam (two phase flow) releasing energy and causing
vibrational damage and/or erosion of the TSP/baffle plates.

Example of
erosion of
Tube
Support
Plates out in
the
condensing
zone.

Erosion of Tube Support Plate / Baffle in Drain Cooler
Segmented
TSP’s
Drain Cooler
Zone
Eroded
(Shroud
TSP
Removed)
Tubesheet
End
Unit Inverted Drain
Condensing Cooler
Zone End Plate

Actual ET results utilizing general voltage percent analysis technique
Actual removed TSP from Drain Cooler


Remnants of TSP ligaments that can give small but false signals of supports

Feedwater Heater Drain Cooler End Plate
The drain cooler end plate is normally much thicker (2.0” – 4+”) than the tube support plates
(0.625” – 0.750”) and liquid between the tube and hole acts as a meniscus to seal the pressure
differential. In some cases the FWH manufacturer actually expanded the tube at the Drain
Cooler End Plate (DCEP) to get a better seal. Sometimes this could lead to failures from
cracking because the tube could become locked and not allow thermal expansion.
If steam bypasses through the tube hole in the DCEP the tube hole can erode and become
larger allowing more steam bypass. If expansion of the tubes in this location is planned then
Array Probe ECT should be used to determine if the tube holes have an smooth edge or have
been eroded to a knife edge which could lead to future problems.
The expansion process is usually until the tube makes contact and then released so the tube
rebounds slightly so it does not lock in place like a tubesheet expansion thus allowing thermal
expansion.
The following examples illustrate ECT indications that could be developing cracks which were
pre-emptively plugged.

This is an ET
graphic of
Tube # I-3-23
showing a
DCEP with
only a slight
expansion.

This is also
an ET
graphic of
Tube # I-3-23
showing the
expected
expansion
was placed
beyond the
DCEP.

This is an ET
graphic of
Tube # I-29-16
showing a
DCEP with a
distorted
expansion
indicating a
possible
crack.

This is an ET
graphic of
Tube # I-30-5
showing a
DCEP with a
distorted
expansion
indicating a
possible
crack.

This is an ET
graphic of
Tube # I-30-6
showing a
DCEP with a
distorted
expansion
indicating a
possible
crack.

This is a
bobbin probe
ET graphic of
a Tube
showing a
DCEP with a
distorted
expansion
indicating a
possible
crack.

This is an
array probe
ET graphic
of the same
Tube
showing a
DCEP with a
crack
indication.

If there are failures at the DCEP then prior to performing any expansion as a
repair the following is recommended:
1) Determine the condition of the DCEP with an Array ET probe (square or
knife edge).
2) If the hole surface is conducive (not knife edge) then expand the tube
to make a better seal.
3) If vibration / fatigue damage was present prior to expansion, consider
inserting a sleeve which would be seal rolled to the parent tube. This
not only provides a secondary boundary in case the parent tube fails
but also would change the vibrational resonance of the tube with in that
immediate area which may mitigate any vibrational / fatigue problems.

If there are failures at the DCEP then a couple of concerns are:
1) The reported thickness (1.250”) of the DCEP seems rather small to start
with.
2) The T-304SS tube wall thickness (0.028”) also seems rather thin and the
resonance value may be conducive to vibration especially if power
uprates have been performed which could lead to flows above original
design factors.
3) A good understanding of the hole surfaces and diameters needs to be
developed so the tube is not expanded out to a sharp contact edge
and/or the tube wall thickness is not reduced much more to seal an
enlarged tube hole. The actual design / ‘as built’ drawings should have
the original hole diameter listed.

While this is an
example of
erosion of Tube
Support Plates
out in the
condensing
zone this
illustrates how
much area
could need to
be filled with
expansion thus
reducing the
tube wall
thickness
greatly.

Broken or Inadequate Impingement Plates
The area of
plugs along the
top were from
severe steam
erosion next to
the edge of the
impingement
plate.
The red tubes
in the base
were from loose
parts TS
damage.

Broken or Eroded Tie Rods
The tie rods that hold the TSP’s in place during assembly are usually mild
steel threaded rods with nuts welded in placed on both side of the TSP’s.
They help to position the TSP’s square to each other so the tubes can pass
fairly easily through the tube holes. To protect the mild steel rods from the
steam flow they are typically covered with a piece of stainless steel tube. If
conditions are unfavorable during operation the SS covering can erode and
then the mild steel rod will go next. This can allow the TSP to move in some
locations and potentially lock or bind the tubes during thermal expansion.
They can also cause denting when they move.
Broken tie rods have also been found in FWH’s that was attributed to
improper operation. They were heated up to quickly during start-up and
cooled down to quickly during shut-down. This caused tubes to become
locked at the TSP’s and can lead to stress cracks in un-annealed U-bends.

The tie rod is
severely
eroded but
has not
failed at this
time. The
TSP in this
same
location has
been
completely
eroded away.

This is a
wider shot
of the same
location
showing
the eroded
tie rod and
missing
TSP.

Case #1:
One plant found pieces of FM in the inlet pass of the highest pressure FWH and it
was identified as the SS tube covering of the tie rods in the nest lower pressure
heater. They were eroded by steam flow and went out through the drain and then
went through the HP pump and got chopped up into smaller pieces. The pieces
large enough to not go into the tubes were removed from the channel head. Then
ALL tubes were then shot with felt plugs with water and air from the upper (outlet)
pass to ensure there was nothing left in the tubes. This reverse flush is important
because if they stopped at the U-Bend it would be easier to move them from the
opposite direction.
The FWH’s were replaced for power uprates within a couple outages and no tube
damage was identified up to that point.

Case #2:
The plant found pieces of FM in the screen basket on the FWH drain
tank that lead from the second highest pressure FWH. The pieces
were fairly large and were easily identified as the SS tube covering
of the tie rods in the FWH. They were eroded by steam flow and went
out through the drain and were caught in the basket.
This FWH was also previously included in the discussion of TSP
erosion. Before FWH replacement was to be scheduled this unit has
been retired for other reasons.

Stress Corrosion Cracking at Tubesheet Crevice / Roll
While this has been identified in the past the chemistry has
stabilized in the industry and the tubesheets can be expanded to
eliminate the crevices.
There needs to be three conditions present to have SCC in T-304 SS:
1. High residual stresses like at the roll transition.
2. A concentration of Stainless Steel susceptible corrodant like
chloride within the tubesheet crevice.
3. Temperatures that exceed 150°F.

Leaking / Ejecting Plugs and stabilizing Cable
It has been identified that when tubes are plugged they can become
pressurized during operation. It is believed any moisture remaining in the
tube heats up and expands. If the plug is not seated very securely and is
disrupted it can become a projectile as pressure ejects it from the tube.
There have also been instances where the lubricant on the wire rope
stabilizer cable broke down during operation temperature and ignited
during the ejection/decompression process and created a fireball.
It has been reported if there is no actual tube leak prior to plugging that
some plants now puncture the tube to allow the pressure to neutralize.

Review of Design Drawings
Whenever possible the design / ‘as built’ drawings should be reviewed
prior to ECT inspections. Important information can be identified like shell
penetrations and/or damage prone areas that should be included in the
inspection scope. Also other information that can be found on the HX
specification sheet like tube material, diameter, wall thickness (BWG),
length(s), tubesheet/tube support materials and thicknesses, operating
temperatures and pressures, shell & tube side medians. All these items can
become very useful in identifying damage mechanisms and possible
corrective actions. This can also help determine if modification of the
inspection approach or technique needs to be incorporated to identify
potential damage mechanisms.
Also actual design or manufacturing errors can be identified if the ET
results show conditions (# of or spacing of TSP’s, tube lengths…) that do
not match what the drawings / specification sheet document.

Modifying ET Techniques
In some instances the inspection technique should be modified to ensure
the damage mechanisms of concern can be identified during the analysis
process. This can include, but is not limited to; inspection frequency(s)
(kHz), screen sensitivity (higher or lower), probe type (U-bend, array,
MRPC, magnetic saturation). In one instance to help identify a potential
problem a duplicate process channel(s) can be used at a very high or low
screen span to ensure it be seen. In another instance if unexpected
indications are found then a magnetic saturation probe should be used to
discount permeability variations.
An open mind and flexibility in approach are very important to ensure tube
integrity because if you aren’t looking in the right direction you won’t see
it.

As Found Inspection
Performing an ‘as found inspection’ is very important. In the case
illustrated below, the craft opened a FWH and then contacted engineering
that it was opened but that they found items in the channel head. Upon ‘as
found inspection’ it was determined the items were the partition plate
retaining nuts and washers. The utility had taken precautions to prevent
this but the tie wire had failed. A number of the nuts and washers became
FM in the inlet pass channel head. Over time they peened the tube end seal
welds so much that inspection of a very large number of tubes was not
possible. This was because the diameter of the inspection probe would
need to be so small to pass it would not meet industry recommendations or
give meaningful results.
This condition was documented and corrective actions were taken to
prevent this from reoccurring.

As Found Inspection
Partition Plate Nuts & Washers

As Found Inspection

As Found Inspection

As Found Inspection
Peened Seal Welds Restricting ET Probe Passage

As Found Inspection
Peened seal
welds
restricting
ECT probe
identified in
RED.

For More Information
Nuclear Division, ANATEC

Darren Howe Paul Koss
VP, Anatec Project Manager
ET Level III QDA ET Level III QDA
Anatec Anatec
Phone: 949-271-7520 Mobile: 615-512-9077
Mobile: 949-300-2173 [email protected]
[email protected]

Steven Schaefer
Manager, BOP Services
ET Level III QDA, ASNT ET III
Anatec, Nuclear Division
Mobile: 860-885-4695
[email protected]
Digitally signed by Steven Schaefer
DN: cn=Steven Schaefer, o=Curtiss-
Wright, ou=Anatec,
[email protected],
c=US
Date: 2017.01.23 12:32:55 -05'00' http://www.curtisswright.com/

Common Problems in Feedwater Heaters: Presented To FSRUG January 19, 2017

Uploaded by

Copyright:

Available Formats

Common Problems in Feedwater Heaters: Presented To FSRUG January 19, 2017

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Common Problems in Feedwater Heaters: Presented To FSRUG January 19, 2017

Uploaded by

Copyright:

Available Formats

Common Problems in

1 | January 23, 2017 | © 2014 Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

2 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

PROPRIETARY and CONFIDENTIAL

4 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

5 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

6 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

7 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

8 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

9 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

10 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

11 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

12 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

13 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

14 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

15 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

16 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

They were severed

17 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

18 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

19 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

20 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

21 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

22 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

23 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

24 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

25 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

PROPRIETARY and CONFIDENTIAL

27 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

28 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

29 | January 23, 2017 | © Curtiss-Wright

PROPRIETARY and CONFIDENTIAL

30 | January 23, 2017 | © Curtiss-Wright