Tccnologías Condensador PDF

Condenser Technology Conference
Proceedings
10937790
10937790
1010322
Proceedings, August 2005
EPRI Project Manager

J. Stallings
ELECTRIC POWER RESEARCH INSTITUTE

3420 Hillview Avenue, Palo Alto, California 94304-1395 • PO Box 10412, Palo Alto, California 94303-0813 • USA
800.313.3774 • 650.855.2121 • askepri@epri.com • www.epri.com
10937790
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN
ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH
INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE
ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM:
(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I)

WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR
SIMILAR ITEM DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS
FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR
INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL
PROPERTY, OR (III) THAT THIS DOCUMENT IS SUITABLE TO ANY PARTICULAR USER'S
CIRCUMSTANCE; OR
(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER

(INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE
HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR
SELECTION OR USE OF THIS DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD,
PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT.
ORGANIZATION(S) THAT PREPARED THIS DOCUMENT
Electric Power Research Institute (EPRI)
ORDERING INFORMATION
Requests for copies of this report should be directed to EPRI Orders and Conferences, 1355 Willow
Way, Suite 278, Concord, CA 94520, (800) 313-3774, press 2 or internally x5379, (925) 609-9169,
(925) 609-1310 (fax).
Electric Power Research Institute and EPRI are registered service marks of the Electric Power
Research Institute, Inc.
Copyright © 2005 Electric Power Research Institute, Inc. All rights reserved.
10937790
CITATIONS
This report was prepared by
Electric Power Research Institute (EPRI)

3420 Hillview Avenue
Palo Alto, CA 94304
Principal Investigator
J. Stallings
This report describes research sponsored by the EPRI.
The report is a corporate document that should be cited in the literature in the following manner:
Condenser Technology Conference. EPRI, Palo Alto, CA: 2005. 1010322.
10937790
3
10937790
PRODUCT DESCRIPTION
This report documents the proceedings of EPRI’s 2005 Condenser Technology Conference.
Background
A large body of information exists on damage in condenser tubes and related condenser
components such as tubesheets and waterboxes. Although technical understanding of damage
mechanisms is generally good, past efforts often have not emphasized the practical, action-
oriented steps needed to effectively deal with the various forms of damage. These proceedings
were organized to help utility users identify, resolve, and prevent tube failure.
Objectives
• To provide a forum to exchange knowledge and experience.
• To identify major problems.
• To establish a clear understanding of priorities for additional R&D to improve condenser
technology.
Approach
The conference was held in San Diego, California, and spanned two days, 31 August to 1
September, 2005. Participants included representatives from vendors, consultants, utilities, and
EPRI experts.
Key Points
Each of the conference’s two days included a plenary session. Plenary presentations for day one
were “Condenser Tube Failures—Theory and Practices: An EPRI Manual,” “Improving The
Reliability of a BWR Through a Condenser Condition Assessment,” and “Modification of
Extraction Steam Expansion Joints in Low Pressure Condensers to Improve Plant Reliability,
Availability, and Reduce Maintenance.” Day two’s plenary session topics were “Vibration
Prevention Goes Beyond Support Plate Spacing,” “AmerenUE Labadie Unit 2 Copper-Beryllium
Tube Development of a Cobalt-Phthalocyanine Catalytic Thin Film to Inhibit Tubeside Biofilm
Adherence in Copper-Beryllium Heat Exchanger Tubing,” and “Gray Water Cooling in Surface
Condensers and Heat Exchangers—Trend or Aberration.” Over the two days, multiple
presentations also were made on both performance and maintenance technologies as well as
fouling control. Each day ended with a panel discussion, “Innovations in Performance
Technology” on day one and “Innovations in Maintenance Technology” on day two.
5
10937790
Keywords
Condensers
Maintenance
Fouling
Corrosion
Failure Analysis
Performance
6
10937790
ABSTRACT
This report documents the proceedings of EPRI’s Condenser Technology Conference, held 31
August – 1 September 2005 in San Diego, California. This meeting, held every three years, is
designed and organized to provide a forum for the exchange of knowledge and experience, to
identify major problems, and to determine priorities for additional R&D. The conference
attracted participants from throughout the industry, including vendors, consultants, utilities, and
EPRI experts. Examples of the six papers presented over two days in the plenary sessions
included “Condenser Tube Failures—Theory and Practices: An EPRI Manual” and “Gray Water
Cooling in Surface Condensers and Heat Exchangers—Trend or Aberration.” Multiple
presentations in separate sessions also were made on fouling control and performance and
maintenance technologies. Each conference day ended with panel discussions exploring a variety
of topics, including but not limited to performance monitoring, leak detection, and condition
assessment.
7
10937790
10937790
AGENDA
CONDENSER TECHNOLOGY CONFERENCE
AUGUST 31-SEPTEMBER 1, 2005
WEDNESDAY, AUGUST 31, 2005

7:00am Registration and Continental Breakfast
8:00am Welcome
Conference Chair: John Tsou, Consultant
EPRI Project Manager: Jeff Stallings, EPRI
Session 1 Plenary Session I

8:30am Condenser Tube Failures – Theory and Practices: An EPRI Manual
Kevin J. Shields, Barry C. Syrett, EPRI
Warren P. McNaughton, Cornice Engineering Inc.
9:00am Improving The Reliability of A BWR Through A Condenser

Condition Assessment
D. Burns, R. Stevens, J. M. Burns, J. S. Burns, Burns Engineering Services
E. Sorenson, Duane Arnold Energy Center
B. Lawrence, Nuclear Management Co. LLC
9:30am Modification of Extraction Steam Expansion Joints in Low Pressure Condensers

to Improve Plant Reliability, Availability, and Reduce Maintenance
Anne M. Edgley, Steven A. Lopez, Kenneth F. Porter, Arizona Public Service
10:00am Break and Exhibits
9
10937790
Session 2 Performance Technology
10:30am On A Predictive Condenser Performance Model
Joseph W. Harpster, Intek, Inc.
11:00am A Study of the Effect of Plant Variables On The Performance of The Pilgrim
Condenser
Norman Rhodes, Charles D. Hardy, Heat Exchanger Systems, Inc.
11:30am The Cost of Dirty Condenser Tubes

Remi Pullen, Technos, Alstom Group
Fred Lherminier, Technos Consultant
Noon Lunch and Exhibits
Session 3 Maintenance Technology I

1:00pm Eddy Current Application for Balance-of-Plant Heat Exchangers
Nathan Muthu, EPRI
1:30pm Remote-Field (RFT) Inspection Techniques to Characterize and Size OD

Cracks at TSP Location for Thin Walled High Performance Ferritic Stainless
Steel Tubing
Nathan Muthu, EPRI
2:00pm Activities, Specialized Tools, and Procedures for Retubing Large Steam Surface
Condensers
Edward E. Overmyer, James W. Smith, Mark Tagliamonte, Retubeco, Inc.
2:30pm Improving Condenser Efficiency With Innovative Scale Removal System

Technology
Jon T. Hansen, Omaha Public Power District
George E. Saxon, Jr., Conco Systems, Inc.
3:00pm Break and Exhibits
10
10937790
3:30pm Panel Discussion: Innovations in Performance Technology
Topics to be explored:
1. Performance Monitoring
2. Performance Testing, Calculation, Trending
3. Fouling Monitor
4. Field Modification
5:00pm Hosted Reception and Exhibits
THURSDAY, SEPTEMBER 1, 2005

7:00am Continental Breakfast
Session 4 Fouling Control

8:00am Condenser Biofouling at Entergy’s Gerald Andrus Power Plant Diagnosis and
Control Strategy
Darrell R. Coker, Applied Specialties, Inc.
Paula Weaver, Entergy
8:30am On-Line Automatic Tube Cleaning System and On-Line Self Flushing Debris
Filter
Kaveh Someah, Brackett Green USA, Inc.
9:00am The Practical Application And Innovation of Cleaning Technology for

Condensers
Andy Howell, Xcel Energy
George E. Saxon, Jr., Conco Systems, Inc.
9:30am Technology for Cleaning The External Surface of Air-Cooled Condensers

George E. Saxon, Jr., Gary Fischer, Conco Systems, Inc.
10:00am Break
11
10937790
Session 5 Plenary Session II
10:30am Vibration Prevention Goes Beyond Support Plate Spacing
David H. Cooley, Edward K. Yost, Alstom Power
11:00am AmerenUE Labadie Unit 2 Copper-Beryllium Tube Development of a Cobalt-

Phthalocyanine Catalytic Thin Film To Inhibit Tubeside Biofilm Adherence in
Copper-Beryllium Heat Exchanger Tubing
Glenn Tiffin, AmerenUE
Bill Nielsen, Bob Kusner, BrushWellman, Inc.
Maan Jawad, Global Engineering & Technology, LLC
Wayne Mueller, Ameren
11:30am Gray Water Cooling in Surface Condensers and Heat Exchangers – Trend or
Aberration
Dennis J. Schumerth, VALTIMET, Inc.
Noon Lunch
Session 6 Maintenance Technology II

1:00pm Condenser Tube Plugs
Mark Meltzer, PSEG Nuclear LLC
1:25pm New Developments In Condenser Tube Coatings

Michael J. Horn, James E. Mitchell, Plastocor, Inc.
1:50pm Condenser Renovation and the Application of Epoxy Coating on Brass Tubes
Jean F. Brunel, EDF
2:15pm Heat Exchanger Tube Structural Repair Sleeves

Bruce W. Schafer, Framatome ANP
James H. Batton, Duke Energy
12
10937790
2:40pm Long Term Restoration to Address Condenser Tube Inlet End
Erosion/Corrosion
W. H. Stroman, Primary Energy Company
Terry Quinn, CTI Industries, Inc.
3:05pm Break
3:30pm Panel Discussion: Innovations in Maintenance Technology

Topics to be explored:
1. Condition Assessment
2. Leak Detection
3. Tube Plugging. Coating, Sleeving
4. Retubing, Replacement
5. Fouling Control
5:00pm Adjourn
13
10937790
10937790
CONTENTS
WEDNESDAY, AUGUST 31, 2005

Session 1 Plenary Session I............................................................................................... 19
Condenser Tube Failures – Theory and Practices: An EPRI Manual

Kevin J. Shields, Barry C. Syrett, EPRI; Warren P. McNaughton, Cornice
Engineering Inc. .............................................................................................................. 21
Improving The Reliability of A BWR Through A Condenser Condition Assessment
D. Burns, R. Stevens, J. M. Burns, J. S. Burns, Burns Engineering Services; E.
Sorenson, Duane Arnold Energy Center; B. Lawrence, Nuclear Management Co.
LLC.................................................................................................................................. 37
Modification of Extraction Steam Expansion Joints in Low Pressure Condensers to
Improve Plant Reliability, Availability, and Reduce Maintenance
Anne M. Edgley, Steven A. Lopez, Kenneth F. Porter, Arizona Public Service .............. 51
Session 2 Performance Technology ................................................................................. 71
On A Predictive Condenser Performance Model

Joseph W. Harpster, Intek, Inc........................................................................................ 73
A Study of the Effect of Plant Variables On The Performance of The Pilgrim
Condenser
Norman Rhodes, Charles D. Hardy, Heat Exchanger Systems, Inc. .............................. 83
The Cost of Dirty Condenser Tubes
Remi Pullen, Technos, Alstom Group; Fred Lherminier, Technos Consultant................ 97
Session 3 Maintenance Technology I.............................................................................. 107
Eddy Current Application for Balance-of-Plant Heat Exchangers

Nathan Muthu, EPRI ..................................................................................................... 109
Remote-Field (RFT) Inspection Techniques to Characterize and Size OD Cracks
at TSP Location for Thin Walled High Performance Ferritic Stainless Steel Tubing
Nathan Muthu, EPRI ..................................................................................................... 151
15
10937790
Activities, Specialized Tools, and Procedures for Retubing Large Steam Surface
Condensers
Edward E. Overmyer, James W. Smith, Mark Tagliamonte, Retubeco, Inc.................. 181
Improving Condenser Efficiency With Innovative Scale Removal System
Technology
Jon T. Hansen, Omaha Public Power District; George E. Saxon, Jr., Conco
Systems, Inc.................................................................................................................. 219
THURSDAY, SEPTEMBER 1, 2005
Session 4 Fouling Control................................................................................................ 229
Condenser Biofouling at Entergy’s Gerald Andrus Power Plant Diagnosis and

Control Strategy
Darrell R. Coker, Applied Specialties, Inc.; Paula Weaver, Entergy ............................. 231
On-Line Automatic Tube Cleaning System and On-Line Self Flushing Debris Filter
Kaveh Someah, Brackett Green USA, Inc. ................................................................... 239
The Practical Application And Innovation of Cleaning Technology for Condensers
Andy Howell, Xcel Energy; George E. Saxon, Jr., Conco Systems, Inc. ...................... 251
Technology for Cleaning The External Surface of Air-Cooled Condensers
George E. Saxon, Jr., Gary Fischer, Conco Systems, Inc. ........................................... 269
Session 5 Plenary Session II............................................................................................ 295
Vibration Prevention Goes Beyond Support Plate Spacing

David H. Cooley, Edward K. Yost, Alstom Power ...................................................... 297
AmerenUE Labadie Unit 2 Copper-Beryllium Tube Development of a Cobalt-
Phthalocyanine Catalytic Thin Film To Inhibit Tubeside Biofilm Adherence in
Glenn Tiffin, AmerenUE; Bill Nielsen, Bob Kusner, BrushWellman, Inc.; Maan
Jawad, Global Engineering & Technology, LLC; Wayne Mueller, Ameren ................... 307
Gray Water Cooling in Surface Condensers and Heat Exchangers – Trend or
Aberration
Dennis J. Schumerth, VALTIMET, Inc. ......................................................................... 347
Session 6 Maintenance Technology II............................................................................. 399
Condenser Tube Plugs

Mark Meltzer, PSEG Nuclear LLC ................................................................................ 401
New Developments In Condenser Tube Coatings
Michael J. Horn, James E. Mitchell, Plastocor, Inc. ...................................................... 429
16
10937790
Condenser Renovation and the Application of Epoxy Coating on Brass Tubes
Jean F. Brunel, EDF...................................................................................................... 437
Heat Exchanger Tube Structural Repair Sleeves
Bruce W. Schafer, Framatome ANP; James H. Batton, Duke Energy .......................... 439
Long Term Restoration to Address Condenser Tube Inlet End Erosion/Corrosion
W. H. Stroman, Primary Energy Company; Terry Quinn, CTI Industries, Inc. .............. 451
17
10937790
10937790
1
SESSION 1: PLENARY SESSION I
10937790
19
10937790
Condenser Tube Failures – Theory and Practice: An EPRI Manual
______________________________________________________________________
Condenser Tube Failures - Theory and Practice:

An EPRI Manual
Kevin J. Shields, EPRI

Barry C. Syrett, EPRI
Warren P. McNaughton, Cornice Engineering, Inc.
Abstract
There is an overwhelming body of information pertaining to damage in condenser

tubes and related condenser components such as tubesheets and waterboxes.
Unfortunately, there has been no single source of reference for the utility user that
details pertinent information on tube failure identification, resolution and prevention.
Further, although the technical understanding of damage mechanisms is generally
good, past efforts often have not emphasized the practical, action-oriented steps
needed to effectively deal with the various forms of damage. The authors are
preparing a manual to address these shortcomings; this paper describes that effort
and the results. The manual identifies all known damage mechanisms that affect
condenser tubes. For each major damage type, there is a discussion of the features
of damage, susceptible materials, susceptible environments, most common
locations, mechanism, root causes, methods of determining the extent of damage,
short- and long-term actions to prevent or mitigate repeat failures, and ramifications
to the unit. In addition there is extensive supporting technical information that will
help the user implement recommended actions. The manual is intended to help
users to (i) anticipate and prevent damage from occurring, (ii) identify, evaluate, and
correct tube leaks and damage if they do occur, and (iii) to initiate actions to help
avoid additional failures.
10937790
21
______________________________________________________________________
Introduction
The demands placed on the condensers of utility generating units are significant.
Functionally, a condenser must condenser several million kilograms/pounds per
hour of wet (~ 10% moisture) steam at low temperatures (typically ~ 38°C or 100°F)
and low absolute pressures (~ 6.8 kPa (1 psia)). It must degasify condensate to the
ppb level. These tasks must be done while also:
• Serving as an impervious barrier between steam/condensate and circulating

water.
• Permitting only limited air in-leakage.
• Contributing minimal corrosion products to the condensate in a “hostile”
environment that is aerated, wet and operates at high fluid velocities.
Over the past forty years, much technical work has been undertaken to understand
the various forms of damage that can afflict condenser tubes. Although there have
been some good publications on the numerous topics that pertain to condensers,
there has not been a single source of information, primarily organized according to
the fundamental damage mechanisms that can lead to condenser tube damage and
leaks. In recognition of the importance of condensers in generating plants and the
needs of end users, the primary objective of EPRI’s Condenser Tube Failure (CTF)
Manual is to provide a compilation of the most recent knowledge about how to
identify condenser tube damage, determine its root cause, and how to apply
immediate solutions and longer term strategies to prevent their reoccurrence.
Additional objectives are:
• To provide a single reference of information about condenser tube failure

mechanisms and the factors that influence them.
• To provide direct, specific actions to be taken if a condenser tube failure has

occurred and provide information about how to anticipate damage that may
be occurring prior to a tube leak.
• To provide sufficient background information so that the user can understand

why the actions are required.
Impact of Condenser Tube Damage/Failures
The impacts of condenser tube leaks and damage on generating units are extensive.
These include effects on cycle chemistry, effects on steam generators and boilers,
effects on the turbine, and effects on plant discharges.
Effects of condenser tube leaks and damage on cycle chemistry. The ingress
of cooling water into the steamside of the condenser will result in its subsequent
distribution into the water-steam cycle. Carried throughout the cycle, there is a
significant potential for damage to the steam generator or boiler, and turbine.
22 10937790
______________________________________________________________________
Effects of condenser tube leaks and damage on steam generators. Steam

generators in both nuclear and fossil units are typically the components that are
most susceptible to damage by contaminants from condenser leakage. In some
cases degradation of steam generator materials has progressed to the point where
replacement of the steam generator has been required. A number of studies have
demonstrated the incentive to minimize ingress of impurities to the feedwater by
leakage of cooling water and/or air into the condenser since such ingress is often the
root cause of damage.
Effects of condenser tube leaks and damage on the turbine. Stress corrosion
cracking and corrosion fatigue of turbine blades and discs are very significant
industry problems for which loss of condenser integrity is considered to be the root
cause of the problems.
Effects of condenser tube damage on plant discharge. Excessive waterside

corrosion of copper alloy tubes can lead to significant levels of copper in the
discharge of once-through systems.
Other effects of condenser tube leaks. Condenser leaks can also increase the
frequency of regeneration in condensate polishers and makeup demineralizers with
the corresponding increase in costs and manpower. Condenser inleakage may also
necessitate increased steam generator blowdown to protect the steam generator
and turbine.
As a result of these significant plant impacts, it is important that condenser tube

damage and leaks be understood and controlled in both fossil-fueled and nuclear-
fueled units.
Key Perspectives for Approaching CTFs
The following precepts represent some of the key philosophical viewpoints taken in
the CTF Manual.
• It is paramount to understand and apply the concept that a clear

understanding of both mechanism + root cause ⇒ an effective and
permanent solution. It is critically important to identify the damage
mechanism and its root cause in order to determine the appropriate solution
to the problem. Applying an immediate or long term solution without
understanding the root cause of the failure is to invite repeated failures.
• Many corrosion modes in the condenser can be prevented or minimized

by giving proper attention to only two aspects of operation and
maintenance. First, if air is eliminated from the condensing steam such
damage mechanisms as condensate corrosion and steamside stress
corrosion cracking of copper alloy tubes can be prevented or minimized. This
is also crucial to unit efficiency and performance. Further, it is necessary to
control air in-leakage to optimize the feedwater chemistry. Second, if
condenser tubes are kept clean and free of deposits, debris, and biofouling on
10937790
23
______________________________________________________________________
the water side, damage mechanisms such as sulfide attack, dealloying,

erosion-corrosion, crevice corrosion, pitting, and waterside stress corrosion
can be prevented or minimized. Again, such action also contributes directly
to unit performance.
• It should be understood that properly selected and optimized cycle

chemistry should not lead to any CTF issues if the condenser design
and operation is proper. The organization should therefore have a goal
that there will be no steam generator tube failures nor turbine damage
or failures as a result of condenser tube failures. This goal is clearly
achievable but does require that any condenser leaks that do occur are
identified, and responded to quickly.
• It is important that each incident (or occurrence) of condenser tube

damage be understood. After initially responding to a condenser leak with
plugging or other means to get the unit back on line, implementation of a
measured (rigorous) program at the next scheduled outage is critical in order
to make a more complete assessment of the extent of affected areas, to
identify the mechanism and perform root cause analysis, and to define and
apply long term actions needed to avoid future damage. To continue to plug
tubes without understanding the cause of the problem is to risk significant
damage throughout the unit. All unit outages should be regarded as potential
opportunities to learn more about the condition of the condenser.
• The necessary technical understanding and the solutions needed to

mitigate outbreaks of condenser tube damage are known and are
available; it is important to make sure that the information is systematically
applied to outbreaks of condenser tube damage and failures, and to
prevention of future damage.
• Management needs to take a proactive stance that recognizes the

importance of CTF on unit availability and reliability and enables the
staff to design and implement a systematic approach to CTF prevention
and mitigation.
Damage Mechanisms
Table 1 contains a list of the damage mechanisms affecting water-cooled

condensers covered in the manual. For each damage mechanism, the following
information is provided:
24 10937790
______________________________________________________________________
Introduction
1.0 Nature of Damage
1.1 Features of damage
1.2 Susceptible locations, materials, and environments
1.2.1 Susceptible locations in the condenser
1.2.2 Susceptible materials
1.2.3 Susceptible condenser environments
1.3 Indications that this damage is active without inspecting the condenser.
1.4 Unit precursors indicating this damage mechanism may become active
2.0 Mechanism
3.0 Root Causes
4.0 Determining the Extent of Damage
5.0 Repair and Immediate Actions
6.0 Long Term Actions and the Prevention of Repeat Failures
7.0 Possible Ramifications or Implications to Other Parts of Unit
8.0 Case Studies
Roadmap for Identifying, Evaluating and Anticipating Condenser Tube

Failures
The manual is structured around a roadmap for use in identifying, evaluating and
anticipating condenser tube failures. An investigation of condenser tube damage
and the corresponding set of actions to be taken can be triggered by four means: (i)
detection of a cooling water leak, (ii) damage being found during inspection, (iii) unit
or condenser performance degradation, or (iv) finding of a “precursor” other than
performance degradation. Figure 1 shows a portion of the roadmap pertaining to
actions to be taken when a cooling water leak is detected or damage is found during
inspection.
The most common path taken for evaluating condenser tube damage starts when a
cooling water leak has been detected, usually by monitoring of water or steam
chemistry. Once a leak of any significance has been detected, it is critical that the
unit be shutdown and the leak(s) located. Under most circumstances, there will be
tremendous pressure to quickly plug the affected condenser tubes and restart the
unit. The second box in the center of Figure 1 indicates some of the actions that
might be taken during the brief plugging outage. In addition to identifying leaking
tubes and plugging them, a preliminary inspection can provide useful information for
the root cause analysis effort. It may be possible to remove samples of suspicious
deposits and biofilms; however, usually, pulling samples of damaged tubes will not
be possible during this short repair outage and will have to wait until the next
scheduled outage. It is important that plugging criteria (number of tubes that can be
plugged) be established a priori so that any repair outages can proceed quickly and
the unit can be returned to service.
Carefully maintained records of the number and locations all plugged tubes need to
be kept. This information will be valuable for a number of reasons, not the least of
which is to serve as input in the subsequent failure investigation.
10937790
25
______________________________________________________________________
During the next scheduled outage, it will be important to continue the failure
investigation and to obtain needed information that could not be gathered during the
repair outage. Samples of damaged tubes may also be obtained at this time for use
in the failure investigation.
Once damage has been found, it is important that the steps in the middle of Figure 1
be followed. If the mechanism is unknown, such as if it appears to be the first
occurrence of this type of damage, or if a number of repeated failures have occurred
and the mechanism is not known, then the screening tables for each damage
mechanism, most commonly affected materials, appearance/ distinguishing
characteristics and most common locations can be consulted. These can provide a
first suggestion toward a specific mechanism discussion where confirmation can be
sought.
On the other hand, if the investigating team has knowledge from past failure that a
particular mechanism is likely occurring, then the appropriate mechanism chapter
can be consulted directly.
Once a mechanism is tentatively identified, the seven Actions shown in the middle
of Figure 1 can be followed.
In Action 1 (Section 1.0 of each major mechanism chapter), the features of the
damage, location, and susceptible materials, discussed in the appropriate chapter,
can be compared with those of the samples removed from service. If the features
seem to match, then the next step (Action 2 and Section 2.0 of each major damage
chapter) is to review the discussion of the mechanism. Do the conditions that
typically lead to this type of damage (environments, stresses, etc.) match those that
have likely been occurring in the condenser?
Once the mechanism is confirmed, a critical step is determining the specific root
cause(s) responsible for this outbreak of damage (Action 3, Section 3.0 in each
mechanism chapter). This is a critical step for a number of reasons. For example,
individual damage mechanisms can have a number of causes. The solutions to be
applied will often depend on the underlying root cause. As a result, if a solution is
chosen without confirming the root cause, it is likely that the problem will continue to
occur. Also, it is imperative that the root cause be determined as some damage
types, notably the corrosion mechanisms, do not occur uniformly over the life of a
tube or other cooling water system component. Thus, for example, corrosion of
condenser tubes that have been in service for 15 years or longer may in fact have
occurred only during the last two years of service. This can make a considerable
difference in evaluating the rate of damage accumulation and forecasting future
problems with the condenser. This is a common mistake: retubing with the same
material but without complete investigation of the damage mechanism and root
cause can lead to short service life of the replacement tubing.
The next step in the investigation of the damage outbreak is to determine the extent
of damage (Action 4 in Figure 1; Section 4.0 in each mechanism chapter). This is
26 10937790
______________________________________________________________________
followed by implementing the appropriate actions (immediate and long term). The
final step will be to determine whether there are implications (“ramifications”) to other
parts of the unit as a result of understanding the conditions in the condenser that
lead to the failure of condenser tubes.
Following the Road Map Stops the Mechanism
Only after the active damage mechanism has been correctly identified is it possible
for the organization to conduct an effective analysis of the responsible root cause or
causes and determine the solution strategies applicable to their mitigation. The
manual includes a table for each individual mechanism that guides the reader
through this process. By proper selection of solutions that address the root causes
identified, it then becomes possible to deactivate or “kill” the mechanism. Failure to
do so invariably leads to future failures by the same mechanism. Use of the road
map approach is intended to eliminate the repeat damage that so often leads to
unacceptable unit availability losses.
As an example, the mechanism of damage by steamside erosion is considered to

illustrate how the process indicated by the Figure 1 road map works.
Damage of this type is typically found on the peripheral tubes that receive the direct
impact of the turbine exhaust and other high energy fluids that are drained, vented,
or dumped into the condenser. An example of damage to tubes at the top of a tube
bundle is shown in Figure 2. The most susceptible materials are the brasses and
90-10 Cu-Ni; with improved performance for 70-30 Cu-Ni. Greater resistance is
obtained from titanium and stainless steel, although no materials are immune.
Damage will only occur on the tube surface facing the flow. Early on, impingement
damage will appear as a polishing of the affected surface. There may be a color
change and/or a dulling of the surface appearance in copper alloy materials. As
damage progresses, the surface becomes increasingly roughened as material
removal increases. Eventually accumulating damage will lead to perforation of the
tube wall or other affected surface. Figure 3 shows the appearance of steam
impingement on 70-30 Cu-Ni tubes. Figure 4 shows steamside erosion of a titanium
tube indicating the roughened surface appearance. Note the absence of erosion at
the position of the support plate in the middle of the photograph.
In ductile materials, a number of responses can occur to repeated liquid droplet

impact. If the impact is intense enough, a single crater can form. For less intense
impacts, microstructural changes occur including deformation, such as by twinning
and work hardening, and the formation of subsurface transgranular fatigue (or
fatigue-like) cracks. The cracks eventually link allowing pieces of material to be
removed. The damage is often uneven on the affected surface as a result of the
random nature of the droplet impacts. Surfaces with fairly advanced damage will
typically display sharp peaks and pits.
Tube damage that matches these features will satisfy the question asked in Action 1
of the roadmap shown in Figure 1, that is, “Do Features Match Those for the Chosen
10937790
27
______________________________________________________________________
Mechanism?”. The next step in evaluating the damage is to confirm that the
damage is consistent with the mechanism of erosion. For example, given a location
on the peripheral tubes and at locations subject to high velocity impingement would
tend to support the thesis that steamside erosion is indeed the correct mechanism.
Table 2, which appears in the manual chapter on steamside erosion, shows the
appropriate table of root causes, steps to be taken to confirm the root cause, and
immediate and long-term actions to prevent additional damage. From inspection of
this table, the user quickly learns that there are five possible root causes that allow
the mechanism to become active, and is provided with confirmation steps needed to
assess their relevancy in individual plants. Taking these steps completes Action 3:
Determine Root Cause.
For example, root cause number 3 is “Excessive steam flows during steam dumps”.
Problems with admissions of fluids may result from:
• The admission fluid conditions may be misunderstood or

underestimated. Under actual operating conditions, the flow, pressure,
and/or temperature may exceed calculated values used for design.
• Inadvertent mixing of steam and water admissions. Steam dispersion

within the condenser may cause serious damage if it is designed or located
so that it picks up water before impinging on internals.
• Improper location of admissions. An incorrect location for the fluid

admission may also result in internal damage. Space limitations and late
design changes may also cause the selection of a poor or improper location.
• Improper operation of admissions. Examples are steam and water dumps

that, although designed for use during startup, are left open too long or are
inadvertently actuated by valve failure, or other means.
• Modifications of the original design. Additional admissions, as well as

replacement of original parts with different materials, have been known to
cause equipment failures.
The industry has recently been experiencing condenser failures associated with
steam dump operation in combined cycle power plants. In contrast to the relatively
infrequent and short-term dumps that were the norm in fossil power plants
historically, in combined cycle units, new modes of operation include1: multiple starts
per year (with steam dumps on startup), multiple HRSG and hence multiple steam
dump lines of varying pressure, steam dump flows up to 180-200% of design flow
and steam dump systems that can operate continuously for hours, if not days or
weeks. This has led to problems with both fatigue and erosion.
To fully characterize the problem, a more complete inspection should be conducted

(Action 4: Determine Extent of Damage) to identify all areas subject to damage by
the mechanism.
28 10937790
______________________________________________________________________
Once the mechanism has been confirmed, the root cause(s) identified, and the
extent of damage determined, the appropriate immediate and longer-term actions
can be taken (Actions 5 and 6 in Figure 1). A list of appropriate actions have been
included in the manual for all major mechanisms, such as listed in Table 2 in for
steamside erosion.
The final aspect of resolution of damage is to determine if there have been any
additional ramifications to the unit as a consequence of this outbreak of damage
(Action 7: Determine Possible Ramifications/Ancillary Unit Problems).
Summary
Despite the significant demands placed on the condenser and exacting penalties for
condenser leaks, the condenser often does not get the attention it deserves. The
Condenser Tube Failure Manual pulls together forty years of research into
condenser tube failure mechanisms. An overall roadmap approach to the
assessment of condenser tube problems has been developed. Specific detail on all
known waterside and steamside condenser damage types has been organized so
that the reader can diagnosis the mechanism, pinpoint the root cause, and apply the
appropriate solutions to prevent or mitigate against future problems.
References
1. Cooley, D.H., “Surface Condensers, Steam Dump & Design Reliability

Considerations”, in Condenser Technology: Seminar and Conference, EPRI, Palo
Alto, CA: September, 2002. 1004116, pp. 1-29 through 1-36.
2. Recommended Practices for Operating and Maintaining Steam Surface

Condensers. EPRI, Palo Alto, CA: 1987. CS-5235.
3. Tavast, J., “Steam Side Droplet Erosion of Titanium Tubes”, in Proceedings:

Condenser Technology Symposium, EPRI, Palo Alto, CA: 1988. CS-5942-SR, pp.
7-1 through 7-17.
10937790
29
______________________________________________________________________
Damage Cooling Water Leak Detected

Found During
Routine
Inspection
To Get the Unit Back on Line:

Preliminary Inspection
Identify Leaking Tubes;
Plug Affected Tubes;
Pull Samples of Deposits, Biofilm, and
(if possible) Damaged Tubes
During Current Outage or at Next Scheduled Outage:

Continue Investigation of Damage
Mechanism Mechanism
Unknown Known;
Verification Required
(includes repeat
failures)
Compare Locations, Go to Appropriate Chapter

Susceptible Materials in Volume 2.
And Appearance of Follow Actions.
Damage to
Tentatively Identify
Candidate Mechanism(s)
Action 1: Do Features
of Damage Match
Those for This
N Mechanism?
Y
Action 2: Determine
(Confirm) Mechanism
Action 3: Determine Root

Cause(s)
Action 4: Determine Extent

of Damage or Affected
Areas
Action 5: Implement
Repairs; Immediate
Solutions and Actions
Action 6:
Implement Long-Term
Solutions
Action 7: Determine
Possible Ramifications/
Ancillary Unit Problems
Figure 1. Roadmap of actions for identifying, evaluating and anticipating condenser

tube failures. This route is taken when damage is found during inspection.
30 10937790
______________________________________________________________________
Figure 2. Impingement erosion failure of 90-10 Cu-Ni tubes at the top of a

bundle.
Source: Reference 2
Figure 3. Impingement erosion of 70-30 Cu-Ni condenser tubes showing

typical roughened surface appearance of damage.
Source: George Moller, Consultant
10937790
31
______________________________________________________________________
Figure 4. Steamside erosion damage in a titanium tube. The absence of

erosion in the middle of the tube was caused by the presence of a support
plate.
Source: Reference 3
32 10937790
______________________________________________________________________
Table 1
Water-Cooled Condenser Tube Damage Mechanisms
Waterside Steamside
• Erosion-corrosion • Steam impingement
• Sulfide attack • Impact damage (foreign object)
• Pitting • Condensate corrosion (ammonia grooving)
• Crevice corrosion • Stress corrosion cracking
• Dealloying • Vibration-induced forms of damage (impact,
• Microbiologically influenced corrosion fatigue, corrosion fatigue, fretting, fretting
• Galvanic corrosion fatigue, cavitation)
• Stress corrosion cracking
• Hydriding damage
• Hydrogen embrittlement cracking
• Cleaning damage
10937790
33
______________________________________________________________________
Table 2
Major Root Causes of Steamside Erosion, Confirmation and Corrective Actions
Root Cause Actions to Confirm Immediate Actions and Longer Term Actions and
Solutions Prevention of Repeat Damage
For all root causes • Repair of damaged steam
flow control devices (baffles,
diffusers, gratings).
• Replace damaged
protective devices such as
clip-on protectors.
• Plug affected tubes.
• If feasible, modify operation
to reduce steam velocities.
1. Locally high (i) Determine susceptible locations • See actions for all root • Redesign and add new baffles
steam flows and potential sources of high causes listed above. or diffusers.
velocity steam. • Reduce flow velocities at inlet
(ii) Perform visual examination to from turbine.
correlate damage with anticipated • Change to more resistant
local velocity profiles. material in susceptible locations.
• Add impingement shields.
2. Improper (iii) Evaluate methods used to • See actions for all root • Redesign and add new baffles
design or prevent problems in damaged causes listed above. or diffusers.
fabrication locations and determine whether (a)
the fabrication matches the
expected design, and (b) that the
design is adequate to handle the
actual steam loads. See
discussions in Section 6 of this
chapter and in Chapter 3 about
designs for handling steam
admission to the condenser.
34 10937790
______________________________________________________________________
Table 2
Major Root Causes of Steamside Erosion, Confirmation and Corrective Actions
(continued)
Root Cause Actions to Confirm Immediate Actions and Longer Term Actions and
Solutions Prevention of Repeat Damage
3. Excessive steam (iv) Develop a condenser • See actions for all root • Redesign and add new baffles
flows during steam connection list. causes listed above. or diffusers.
dumps (v) Review the design of steam • Change to more resistant
dumps for connections in affected material in susceptible locations.
areas. • Add impingement shields.
4. Improper operation (vi) Measure or calculate steam • See actions for all root • If operating procedures cannot
or operating procedures flows in affected areas for all causes listed above, plus be changed, it will be necessary
that are producing operating procedures and • Modify operation to to add additional protective
unexpectedly compare the flow rate and their reduce sources of steam structures including, for example,
aggressive steam flows timing to that expected in the impingement. new baffles or diffusers.
design.
5. Increased steam flow (vii) Review uprating design • See actions for all root • Redesign and add new baffles
with unit uprating calculations to evaluate whether causes listed above. or diffusers.
appropriate design changes have • Change to more resistant
been considered to incorporate material in susceptible locations.
increased exhaust flows. • Add impingement shields.
10937790
35
10937790
IMPROVING THE RELIABILITY OF A BWR THROUGH A CONDENSER
CONDITION ASSESSMENT
Authors: Daniel C. Burns Burns Engineering Services, Inc.

John M. Burns P.O. Box 272
Robert L. Stevens Topsfield, MA 01983-272
Jeffrey S. Burns www.BurnsEngr.com
&
Eric Sorenson Duane Arnold Energy Center
Brian Lawrence Nuclear Management Co. LLC
3277 DAEC Road
Palo, IA 52324
10937790
37
Keywords
Condenser, Component Evaluation, Inspection, Reliability, Availability, Condition Assessment, High Energy
Connection Failures, Erosion, Inspection Criteria, Finite Element Analysis (FEA)
Abstract
Few utilities have comprehensive condenser inspection plans in place with relevant component inspection
criteria. Yet, the methodical inspection is a critical tool for monitoring any deterioration of condenser
internals; through a case study, this paper will present such a method. The carefully developed inspection
plan to be described provides useful information in order to develop an effective maintenance program that
can minimize the potential for future operational reliability or availability issues.
Creating an inspection plan however with appropriate criteria is often a difficult task. Much of the
information that has been collected and the subsequent engineering evaluations that are available have
identified when a particular condenser component condition degraded to a point where repair or replacement
was necessary. These are post-mortems- in comparison to projections into the maintenance future. The latter
larger scope, cost-effective improvements would instead be based on a working knowledge of condensers
and the complex interactions that guide their operation. The paper will outline how that bigger picture
condenser inspection plan is tailored to a specific condenser and plant.
This paper will present a case study that tracks a focused condition assessment program for an almost 30
year-old dual-pressure, two shell condenser serving a 614 MW nuclear Boiling Water Reactor (BWR)
station. The paper will highlight the specifics of a number of independent engineering tasks and how a
comprehensive inspection plan was developed to cost-effectively monitor and maintain the unit condition.
Relatively common condenser conditions will also be addressed by the paper: severe erosion on the shell, its
supports and the tube bundle bracing within the vacuum space had previously been identified by the station
as a serious concern. The Finite Element Analysis (FEA) method was employed to determine the failure
criteria of those structural components for use during the inspection. These criteria could not have been as
accurately determined by any other method. The program allowed more latitude in the scheduling of the
required maintenance with a less costly repair plan. The parameters that cause internal failures will be used
in conjunction with the plant ultrasonic testing and routine outage inspections.
The erosion of several high energy drain connections had previously caused tube failures. The paper will
present methods for review of the operational limits of some of these connections. The inspection work
described will also indicate the extent of local damage due to impingement and erosion on and around these
connections. In addition, some of the initial plant shutdowns were attributed to debris lodgment inside the
tubes that produced a tube wall erosion which caused tube leakage. The paper will also outline those specific
areas that require visual examination.
38 10937790
Introduction
In today’s climate of deregulation, it is often difficult to find the maintenance funding and engineering
resources to utilize a condenser maintenance program that employs the inspection as the main provision for
maintaining unit condition. The condenser is however often the largest piece of balance-of-plant equipment
and there are a number of components that require routine inspection and maintenance to keep unit
availability high and condenser operation reliable.
A case study is presented herein which discusses the important reliability contribution of the visual
condenser inspection based on a current, real power plant experience.
Recently, the management of a BWR in Iowa had concerns with maintenance practices that affected the unit
reliability and performance after a pattern of forced outages and degradation of performance over a 10 year
period. The utility management prudently formed a dedicated team for creation of a reliability improvement
program with the main charter of isolating the previous problems and any potential causes of failure. The
team produced a series of evaluations that culminated in a plan of action entitled the “Condenser Reliability
Improvement Plan”. Several of the major findings of that plan will be discussed.
Multi-level reviews of this 100 plus page document were performed by external groups such as the
condenser manufacturer (Foster Wheeler), a large architect-engineering firm (Black & Veatch) and other
third-party groups such as Burns Engineering Services (BES). The end result was one peer reviewed
document that listed many of the best practices in the industry. The thrust of the report was to return unit
performance back to design level and to minimize condenser reliability issues. It was also used to direct the
team’s effort in instituting a maintenance program that was more proactive, instead of reactive. Best
practices would be used to quickly identify and correct any maintenance item that may later become an
underlying cause of failure.
Burns Engineering Services, Inc. was eventually selected to address each major concern or known
deficiency. They also developed an inspection plan with a focus on preventative strategies including the
early onset and identification of component failures. That company further conducted the inspection during
the refueling outage in the Spring of 2005. In order to ensure an effective return on maintenance resources,
outage action items were to be prioritized using engineering guidelines and criteria as opposed to attempting
a wholesale scheduling of repairs.
Pre-Inspection Engineering Evaluations

Shell Bracing
Severe steam erosion and bowing of the lateral shell supports located in the upper extension neck steam
space between the expansion joint and turbine flange of the condenser had been reported by the station. To
minimize the possibility of buckling of these braces and a potential shell failure in the vicinity, the plant has
been replacing these braces individually at each outage. Hence, with regard to this condenser component, the
task was to determine the maximum number of braces that could be replaced at one time and to develop
inspection criteria for when to replace these pipe columns.
The most accurate, reliable method for determining support requirements of the shell during replacement
activities and also for the purpose of developing inspection criteria was to create a precise structural
computer model of the area including the lateral shell supports. Utilizing finite element analysis (FEA), the
10937790
39
resulting deformations and stresses of those lateral shell supports in several conditions during condenser
operation were assessed for the purpose of establishing the maintenance and inspection requirements.
The station had also reported severe erosion in localized areas on the upper surfaces of the diagonal braces.
None of these had exhibited bowing or had been replaced to date. FEA was hence also employed to provide
an inspection criteria and recommendations to repair or replace these severely eroded diagonal supports as is
described later.
Finite Element Analysis Method

One of the most significant advances in the field of computer-aided structural analysis has been the
development of Finite Element Analysis (FEA). Today, many industries routinely use the method to cost-
effectively reduce or eliminate prototype testing and to optimize the cost, weight and strength of their
designs. These industries include the automotive, aircraft, electronics, transportation and many other sectors.
The technique replaces a component by a finite number of small structural elements such as plates, beams,
bricks, etc. The loads and displacements are linked to each element through the application of Hooke’s law
at the nodes and the fundamental equations describing the sums of the forces and moments are written
around them via the computer code. The result is a set of linear equations in each direction for all the nodes.
The systems of equations become very large matrices that can only be practically solved by a computer.
Their aggregate overall reactions such as the deformation, strain and subsequent stress levels approximate
the mechanical behavior of the entire structure.
Continual increases in computer speed, core size, powerful improvements in the graphics and the
development of better, user-friendly software have made FEA a practical engineering method for the
condenser industry [Ref. 1]. Though now the methods are often used for new condenser design because it is
a highly redundant structure, FEA is also an effective engineering tool for the analysis of existing designs.
When properly conducted, FEA dramatically identifies the weaknesses and also the conservative features of
an operating condenser that years ago was the design result of an empirically developed structure. Only with
extensive (and cost-prohibitive) strain gage testing of the detailed condenser structure could the original
manufacturer of this condenser, Foster Wheeler (FW), have had the similar insights into the effectiveness of
his largely empirical structural design.
In this pre-inspection evaluation, the model, the loading, the constraints and their results were produced from
the application of ALGOR FEA software [Ref. 2]. Burns Engineering has been licensed to apply that
software for over seven years and is well versed in its application to all aspects of condensers. The ALGOR
FEA software code complies with NRC recognized QA criteria, specifically 10CFR50, Appendix B and 10
CFR21. The software also utilizes the ASME accepted structural properties of the lateral shell supports and
shell materials.
Finite Element Condenser Component Models

Based on the FW condenser drawings with Bill of Materials, the virtual FEA computer model reproduced all
the major structural components, including a conservatively high estimate of operating loads with other
constraints.
As examples of the model representation, the exhaust opening size and height of the extension neck between
the turbine skirt and condenser expansion joint were reproduced to within ± 1 inch. This was the section
where deformed lateral corner bracing was noticed. The software separately distinguished and computed the
cross-sectional areas, all beam section moduli and moments of inertia for the 12 - 3 inch and the 4 inch
schedule 40 lateral braces. In addition, the extra clamp-bar thickness at the dog-bone expansion joint was
accounted for; closely modeled also were the small bars that make up the central truss. Besides the dead
40 10937790
weight of the steel and the assumed maximum one atmosphere pressure load on the shell during operation,
we conservatively included a dynamic force of the steam at a maximum velocity of Mach 1. The resulting
model contained almost 1000 nodes using small size plate & beam elements to accurately capture the
behavior of that entire section of the condenser. The primary features of the model are displayed in Figure 1
below, the displacement output when in operation with the conservative loadings previously described.
Figure 1 – Deformation Plot of Upper Shell between Expansion Joint & Turbine Flange
To evaluate the effects of erosion on the lateral shell supports, FEA computer simulations with the maximum
operating condenser loading was employed. These established the stresses and deformations of the lateral
shell support and the shell integrity at the original design condition (with no wall loss). Other runs of the
program developed stresses and deformations with the same load cases for uniformly reduced pipe column
wall losses, and the braces removed. They were based on the varying conditions described below:
Case 1: Lateral shell supports with a uniform 1/16 and 1/8 inch steam impingement erosion and a conservative 1 atm.
(14.7 psia) shell loading were assumed along with a uniform pressure distribution caused by a conservatively estimated
Mach 1 steam velocity.
Case 2: All corner lateral shell supports were removed with the unit out of service so that essentially, only the weight
of the shell was included.
In Case 2, the FEA modeling procedure was to sequentially remove the lateral shell supports starting with
those located in each of the four corners. The computer model load simulation was run and then the
simulation was repeated with even fewer braces; ultimately all the braces would be removed if the stresses
and deformations in the beam were determined to support the shell. Note that the upper and lower mid-span
lateral shell supports and the truss across the exhaust opening at the mid-plane of the long side were not
removed in these simulations.
10937790
41
In each case, a comparison of the resulting stresses and deformations on both the braces and the condenser
shell plate was reviewed to determine the sequential impacts of removal of lateral shell supports. The
allowable stresses from the Heat Exchange Institute (HEI) Condenser Standards [Ref. 3] and our team
experience were utilized as the criteria of the reviews. Figure 2 below shows the FEA element model in finer
detail. It shows the element construction with the required connectivity and the numbers of separate beam
elements that were included.
Figure 2 - FEA Model Cut-Away of Turbine-Condenser Flange
High Energy Drains
Engineering reviews were made of the flow conditions for the intermediate energy and high energy
connections. Criteria used during these reviews included both EPRI’s 1982 Report [Ref. 4] that provided a
direction for the admission of high energy fluids into condensers and another guideline based on an industry
connection design review practice. Connection lists were developed and the above design checks were made
to aid in the inspection assessments of the connections. These methods were used to check the existing shell
connections and penetrations in order to determine if their original designs were compatible with current
design standards. As a general rule-of-thumb, the condenser connection steam velocities should held at or
below approximately 600 ft/sec and water velocities should be at or below approximately 10 ft/sec. On that
basis, the majority of connections were determined to be satisfactory, particularly in view of the long service
of the connections on this particular condenser. As a concrete example, the following Table 1 typical very
small portion of the connection review listing defines and estimates each of their characteristics. Then the
pre-inspection engineering evaluation remarks concerning each major connection are indicated. The latter
42 10937790
suggested that during the inspection a more careful examination should be made of the connection in
question: (The connection numbers refer to those on the condenser drawings.).
Table 1- High Energy Drains Connection Review
Conn.# Size Flow Press. Enthalpy - Flow Spec.Vol Volumetric Flow Area Velocity Service
(Inches) (Lb/Hr.) (PSIA) (Btu/Lb.) (lb/sec) (cuft/lb) (cuft/sec) (sqft) ft/sec Frequency
14 18 (4) 1,442,500 65 1,055 400.7 5.7500 2,303.99 1.6230 1,419.6 Intermittent
21 12 4,000,000 15 98 1,111.1 0.0167 18.58 0.6303 29.5 Intermittent
22 24 3,147,600 1.88 92 874.3 0.0162 14.16 2.9500 4.8 Continuous
23 18 300,000 20 1,144 83.3 20.0000 1,666.67 1.6230 1,026.9 Emergency
….Connection 14
The spray pipe orifice area should be approximately 685 sq. in. This pipe has a design orifice flow area of
1,187 sq. in. At this flow area the expected pressure in the pipe would be approximately 38 psia, not the 65
psia as listed. See Spray Header Detail Connection 14 Dwg. 93-1029-3-189.
….Connection 21
The design type and overall size of the connection is satisfactory. Estimated flashing amount of
approximately 111,500 lbm/hr. At this amount of flashing and an enthalpy of 98 BTU/lbm, a spray pipe is
recommended. The pipe water velocity is high. Spray pipe design details are not available.
….Connection 22
The design type and overall size of the connection is satisfactory. Estimated flashing amount of
approximately 69,500 lbm/hr. For this amount of flashing and an enthalpy of 92 BTU/lbm, a spray pipe is
recommended. Spray pipe design details are not available.
Other Engineering Evaluations
To properly address the quantity of potentially buckled tube bundle support plate pipe columns above the
“false bottom plate” that could be replaced, the loading was estimated and a maximum repeating column
spacing was determined by the application of plate deformation estimates [Ref. 6]. The false bottom is
plating beneath the tube bundles that separate the lower storage hotwell and the upper tube bundle hotwell in
this particular multipressure design.
Additionally, a major connection failure along the high pressure (HP) shell wall had caused catastrophic
tube failures during the last operational cycle. But it was also responsible for the blow-out of large sections
of thin walled vertical plates surrounding a sector of the condensate reheat zone tray in that HP shell. The
damaged vertical plates had experienced fatigue cracks and failed due to the high energy steam jets striking
the plates in one section. Though the absence of these plates seemed benign and not worth the cost and
effort of replacement, BES felt early that the plates contributed to the reheat of the large quantity of
relatively cool condensate from the low pressure (LP) shell to approach the temperature of the HP
condensate. A preliminary assessment of this latter aspect of the design was made with the help of some
difficult-to-read manufacturers condenser design shell drawings. It however remained for the visual
examination of the detailed area up-close and in-person to firmly establish the plates would need to be
repaired/replaced.
10937790
43
A weld in the cross-connection seal steam assembly between the LP and HP shells during the operation cycle
had failed and caused a significant air leak that was difficult to find. No effective pre-inspection engineering
evaluations could be conducted on this component, however, it was marked to be carefully inspected during
the inspection itself.
Not relevant to this case study, but there is an additional study often worth considering evaluation when
condenser tube vibration failures have periodically occurred and the tube bundles are not extensively
installed with anti-vibration stakes. Then, a fluid-elastic vibration analysis should be included. In this
instance, the winter performance, the detailed tube layout and the condenser tube support spans are utilized
in an estimate to determine steam velocities at various tube rows that produce an unstable, large amplitude
mechanical vibration. Since this particular tendency to vibrate is usually greater at the upper and side
peripheries of the bundles, during the outage, it would be prudent to inspect susceptible sections of the tube
bundle so identified for any evidence of tube wear or movement.
Inspection Plan
A main objective of a condenser inspection plan is to provide a framework for monitoring and assessing the
physical condition of the condenser. This provides for identification of the cause(s) of deterioration and any
corrective action(s) necessary to provide a suitable repair or for replacement. The condenser inspection
covered all standard structural components, and those previously identified in the Condenser Reliability
Improvement Plan as problem areas that resulted in forced outages. A brief example of the items listed in the
inspection plan is given below in Table 1. Note that this inspection list is tailored specifically to the case
study but it typifies the condenser components that should be inspected during a scheduled outage.
Inspection Procedures
The condenser condition assessment procedure consisted of a visual inspection that was supported by a team
of experienced BES & station personnel. Safety was of course a major focus; but the team then had members
who were familiar with the particular maintenance and operational history of this condenser and also ones
knowledgeable in the design, performance and general condenser industry problems. The inspection was
performed during four sequential day shifts at this BWR nuclear plant, and was completed within one
business week including a day for the required radiation worker safety training. The inspection was
systematically conducted to avoid overlooking or bypassing areas or duplicating efforts. To conduct an
inspection that was satisfactory to everyone, a detailed plan was provided to the station many weeks previous
to the outage to ensure all areas of concern would be included. The list of components and areas to be the
focus of the inspection were based on the previously mentioned Reliability Improvement Report, typical
historic condenser industry problem areas for that vintage of condenser and the several engineering
assessments that are to be described. The Table 2 listing outlines the scope of the inspection plan. Other
condenser inspection recommendations can be found in Ref. 5.
In order to properly document the inspection findings, the team photographed significant deficiencies found.
Photographs are now an important aid to locate and preserve the extent of the deficiency in question for later
reference.
44 10937790
Table 2- Outline Example of Condenser Inspection Plan
Though the listings below are specifically tailored to the case study plant, they typify the scope of a
comprehensive condenser inspection plan.
Waterboxes
The majority of forced outages were from tube leakage. In this case, based on the tube metallurgical analysis, many of the
tube leaks were reported to be a result of lodgement erosion or the presence of large debris in the circulating water system
lodged at the inlet of the tubes. Hence, the team planned to carefully inspect the internals of the waterboxes and especially
the inlet tube ends as follows:
• Check the waterbox lining for physical damage, blistering, material loss and detachment from the waterbox plate.
• Check for indications of waterbox plate material corrosion. This would include rusting, pitting or erosive wear.
• Check for the amount and type of debris accumulation and buildup in the waterboxes. If tube cleaning is planned,
inspect before cleaning and after to identify prevalent type(s) of debris in tubes.
• Check the tube sheet surface condition. Document any significant erosive/corrosive attack.
• Observe and document the amount and type of inlet end tube blockage.
• Inspect the inlet tube ends for erosive wear, fouling and scale buildup. If possible, shine a light or feel down tube
for the presence of a deposit (that would suggest an under-deposit attack) or pitting. Tubes should be marked for
later ECT or other measurement technique.
• Check the condition of the inlet end tube inserts.
• Check the condition of tube plugs and if any plugs are loose.
• Check the condition of instrument, vent and drain connections.
• Check the condition of any sacrificial anodes.
• Check the condition of the waterbox grating.
Condenser Tube Bundle

Check accessible peripheral tubes for the following defects:
1. Erosive wear: Steam-side erosive wear or steam impingement is caused by high velocity steam and connection flows. These
velocities can reach sonic speeds and will erode the tube surfaces. The eroded surfaces will be rough in appearance. Initially
they will look and feel like the surface of fine or medium coarse sandpaper. In its most severe form, erosion will penetrate the
tube wall. There is usually a pattern to the impingement and often a suspected local connection.
2. Fluid-Elastic Vibration: High velocity steam and connection flows striking tubes causes vibration which can result in tube
failures. These are usually circumferential cracks or tube wall damage that occurs at the mid-span between the tube support
plates at the periphery of the bundle. If vibration is sufficiently severe, adjacent tubes may strike each other causing flat or
dented sections where they hit. Vibration can also cause fretting or wearing of the tubes in the support plate holes or cracks at
the tube sheet. Check also the condition of the vibration stakes. Identify any that have structurally failed or slipped off mid-
span. Mark for repair or replacement.
3. Physical Damage: Objects striking the tubes will cause dents and possible punctures. Tools and materials left in the
condenser or turbine exhaust during maintenance work often are the cause. Other objects consist of failed condenser
components such as connection baffles, deflector plates, spargers, basket tips and other collapsed condenser parts. These
objects become missiles during the start up and operation of the unit.
4. External Coatings: Identify if external tube surfaces are heavily coated with magnetite or other type of oxide layer.
5. Locate & Document Significant Tube Defects: All tubes with severe indications are to be marked and plugged.
Internal Shell Components

Check the tube support plates including their supports, bracing and attachment welds. Locate and document any significant
deterioration including erosion, bowing and broken or cracked welds. Erosive wear up to 1/8 inch is not to be considered
significant. Some bowing of the support plates is expected and may have occurred during installation. Support plate bowing of
greater than 1 inch at its center should be investigated. Gently push tubes to identify any excessive movement or signs of wear
at supports. Document all significant deficiencies.
10937790
45
Check the hotwell tube support plate support posts (pipe bracing) for erosive wall loss. Make estimates of the depth of the erosive
wear. Depths or losses of greater than 1/8 inch are considered to be significant. The team will inspect the attachment welds for
cracks and/or deterioration.
Check the location and condition of all internal shell, steam dome and hotwell partition plates, impingement plates/rods and drain
trays. Locate and document any significant deterioration including eroded, broken and cracked plates or drain trays. The team
will mark any items determined critical to the reliable operation and performance of the condenser for repair or replacement.
Inspect the shell in the vicinity of the last stage of the turbine exhaust. It was reported that sections of the steam dome shell wall,
especially up near the last stage bucket exhaust, had experienced severe erosion. A typical 2'x3' or 3'x3' area with a pattern of
erosion was reported which appears to be a little more pronounced in Unit A (LP side). The plant planned on monitoring the
wall thickness with UT(Ultra-Sonic) tests. As a general rule, any severe erosion should be repaired. Usually, a flat, carbon
steel wear plate welded over the immediate area of damage is sufficient. Note that an FEA model of the area can be used
to obtain failure criteria. These inspection criteria can then be applied for visual inspections or shell UT tests.
Inspect the drain tray vertical plates to confirm their function. Provide immediate recommendations for any action to be taken
including repair/replacement.
Check the condenser steam dome (exhaust hood) lateral braces/support struts for erosive wear, bowing and attachment weld
cracking and/or defects. The team will make estimates of both the depth of erosive wear and the bowing where necessary.
Depths or losses of greater than 1/8 in are considered to be significant. Based on an FEA model engineering study of those
braces that preceded the inspection, all horizontal support braces found with wall losses of 1/8 in or greater shall be marked
deficient for repair, replacement or further evaluation.
Check the condition of the crossover equalizing duct welds. Inspect the weld and structure of the crossover duct for corrosion or
cracking. Determine if any components (i.e., seal plate, trough…etc.) appear to enhance the stiffness of the structure and.
identify any measures that would strengthen that crossover section.
Check the condition of the condenser basket tips and their supports.
Inspect the internal arrangement of all high to intermediate energy condenser connection penetrations. This includes spargers,
baffled and deflector plate arrangements including their welds. All previously failed connections are to be included to make
sure no further corrective action is required. If significant damage is observed at the location of any penetration, document it
and conduct a design review on the estimated actual flow conditions to determine an appropriate long-term corrective action.
Significant damage would include the following:
1. Excessive erosive wear to deflector plates, baffles and sparger pipe spray holes.
2. Excessive erosive wear to adjacent plate, structural members and tubes.
3. Cracked deflector plate and baffle welds.
4. Dislodged and missing deflector plates and baffles.
5. Dislodged or cracked sparger piping and cracked or missing piping end plates.
6. Cracked and missing sparger piping supports, brackets and baffles including their attachment welds.
Inspection Results
The following are a few of the digital photos taken during the actual inspection. All the major results and
recommendations were presented to the plant immediately after the inspection so that timely outage work
orders could be written. This verbal report was then followed up by a confirming email within the same day.
The formal inspection report that documented the findings followed later because that is not needed to as
expedited.
46 10937790
Figure 3-Loose Tube Plug Figure4- Missing Section of Extraction Line Lagging
10937790
47
Figure 5- Eroded Lateral Bracing Figure 6- Sparger Connection End-Cap Failure
48 10937790
Conclusions
As a result of the considerations within this paper and the overall objectives of the station, it was concluded
that:
1. Considering the dynamic environment and wear of an older condenser, it is prudent for station
management to focus on condenser maintenance in a proactive manner, in order to have a control of
the condenser issues that could force them into a response mode.
2. The costs associated with developing this proactive essentially inspection plan are minimal when
weighed against the potential for forced outage.
3. Periodically producing a condenser evaluation of a specific unit for a wide industry review creates an
effective basis for improvement and the subsequent inspection and more particularized engineering
reviews.
4. The discussion sections in this paper list detailed areas and components for inspection that should be
reviewed by utilities contemplating a condenser inspection.
5. Engineering evaluations provided guidelines and criteria for determining if observed deficiencies
were critical and if repair or replacement was required.
6. The approach of utilizing a specialized condenser engineering firm teamed with plant personnel
synergistically provided effective condenser evaluation, inspection and repair recommendations to
enable continued reliable operation.
References
1. Burns, J. M.; Burns, D. C.; Burns, S. J.; “Structural Simulations of Condensers Using Finite Elements
Furnish an Effective Field Repair or Design Basis”, EPRI Condenser Technology Conference, 1999.
2. Linear Finite Element Analysis, V16 -November 2004, ALGOR Inc., 150 Beta Drive, Pittsburgh, PA
15238-2932.
3. Heat Exchange Institute, “Standards for Steam Surface Condensers”, 1995 and Addendum 1, 2002.
4. Recommended Guidelines for the Admission of High-Energy Fluids to Steam Surface Condensers,
EPRI CS-2251, 1982.
5. Stevens, R.; Burns, J., et al: “Condenser Inspections: Valuable and Cost-Effective.” EPRI Condenser
Technology Conference, 1990.
6. “Formulas for Stress & Strain”, R. Roark & W. Young, 5th ed., McGraw-Hill, NY, 1982.
10937790
49
10937790
Modification of Extraction Steam Expansion Joints in Low
Pressure Condensers to Improve Plant Reliability,
Availability, and Reduce Maintenance
EPRI Condenser Technology Seminar and Conference

San Diego, CA, August 30-September 1, 2005
Anne M. Edgley (Arizona Public Service Palo Verde System Engineering)
Steven A. Lopez, P.E. (Arizona Public Service Palo Verde Design Engineering)
Kenneth F. Porter, P.E. (Arizona Public Service Palo Verde System Engineering)
Abstract
The Palo Verde Nuclear Generating Station (PVNGS) consists of three essentially
identical approximate 1400 Megawatt Electric Combustion Engineering (CE) System 80
NSSS Units with General Electric Turbine Generators. Each Palo Verde Unit has a
large three shell Westinghouse Condenser and seven stages of feedwater heating, four
low pressure (heaters No. 1, 2, 3, & 4) and three high pressure (heaters No 5, 6, & 7).
The condenser is a multi-pressure, radial-flow, single pass, shell and straight tube,
reheat type heat exchanger. The condenser is floor supported and located beneath the
three low pressure turbines which each exhausts directly to its associated condenser
section. The Palo Verde Units were put into service between 1986 and 1988.
In November of 1999 Palo Verde Nuclear Generating Station (PVNGS) Unit 2

experienced a rapid sodium excursion in one of the main condenser hotwells along with
a drop in Megawatt output and an increase in noise from the condenser. The
engineering staff indicated that an extraction steam expansion joint maybe leaking
based on examining LP heater pressures and temperatures. A decision was made to
continue operation until the next refueling outage. In January 2000 another of the main
condenser hotwells experienced a much larger sodium increase. After an initiated
shutdown, the condenser waterbox was opened for tube inspections and testing. These
activities revealed a single leaking tube in the top of the tube bundle. After breaking
vacuum, a general condenser shell inspection found that large quantities of expansion
joint debris had fallen onto the tube bundles causing a tube leak.
A multi-discipline root cause team investigated the event. This team determined that two
expansion joints that sustained the most damage and their associated piping system
were not designed for the vibration service conditions. The initial failure of either one or
both of these expansion joints resulted in the subsequent steam flow from these failed
joints impinging on the adjacent expansion joints causing collateral damage of nine
additional joints. The root cause team also determined that the preventive maintenance
10937790
51
tasks that performed joint visual inspections were not sufficient to identify and prevent
the identified failures.
The investigation resulted in both short term and long term corrective actions. The
immediate corrective actions involved an inspection of the affected Unit 2 condenser
during shutdown. The identified damaged expansion joints were replaced with similar
designed expansion joints. In Unit 3 during the subsequent outage expansion joints
were replaced with a slightly different design. Long term corrective actions included;
development and installation of a new engineered expansion joint design better
matched for existing operating conditions. Other actions involved changing extraction
steam piping supports to incorporate the new expansion joint design and to help reduce
pipe vibration and the expansion joint visual inspection preventive maintenance tasks
were revised to increase the frequency and scope of visual inspections.
During subsequent inspections additional axial cracking indications in certain higher

pressure expansion joints have been found suggesting that some expansion joints could
be susceptible to pressurization and thermal cycle degradation. Preventative
maintenance inspections continue and have resulted in consideration of installing
external bellows covers. The results of these inspections continue to be evaluated to
determine if additional corrective actions may be required as more operating experience
is gained and system health is accessed. Industry operating experience is also
reviewed to determine applicability to observed PVNGS extraction steam expansion
joint degradation.
Introduction
Each of the Palo Verde Nuclear Generating Station (PVNGS) Condenser shell contains
4 stages of extraction steam for the four stages of low pressure feedwater heaters. The
12th stage LP extraction steam piping contains eight 20” vertical universal expansion
joints that feed into a common ring header. This ring header has four 28” vertical lines
that feed the 1st point heater. The 11th stage LP extraction steam piping contains four
16” lines that feed the 2nd point heater. All four lines contain a 16” vertical universal
expansion joint while two of these lines also include a 16” horizontal expansion joint.
The 9th stage LP extraction steam piping contains two 18” vertical universal expansion
joints that feed into a 28” common header that leaves the condenser shell. PVNGS
added a 28” horizontal universal expansion joint to this common header. Outside the
condenser, the piping continues to a single Bleeder Trip Valve (BTV) and two isolation
valves before returning to the condenser. The piping continues inside the condenser to
a 28” pressure balanced expansion joint before entering the 3rd point heater. The 8th
stage LP extraction steam piping contains two 12” vertical universal expansion joints
that feed into a 18” common header that leaves the condenser shell through an 18”
horizontal universal expansion joint. Outside the condenser, the piping continues
through two Bleeder Trip Valves (BTV) and two isolation valves before returning to the
condenser. The piping continues inside the condenser to an 18” pressure balanced
expansion joint before entering the 4th point heater.
52 10937790
The 1st and 2nd point heaters also include a single pressure balanced expansion joint on
the heater drain lines. The 1st point heater has a 16” drain line while the 2nd point heater
has a 12” drain line. The 3rd and 4th point heater drains used to include expansion joints
also, but they were replaced by hard piping after multiple expansion joint failures were
experienced.
Background
The original PVNGS Extraction Steam Expansion Joints were supplied by one
manufacturer, Expansion Joints division of Process Engineering, Inc. Currently PVNGS
uses 12", 16", 18", 20", and 28" diameter extraction expansion joints supplied by one or
more of five manufacturers;
• Steam Expansion Joints
• Expansion Joints division of Process Engineering (original manufacturer)
• Flexible Metal Hose Company (new manufacturer)
• Expansion Joint Systems, Inc. (new manufacturer)
• Senior Flexonics, Inc. (new manufacturer)
• Drain Line Expansion Joints
• ADSCO Manufacturing Corp. (original manufacturer)
The original supplied expansion joints were designed for a minimum life of 10,000
thermal cycles. Piping and expansion joint vibration have been attributed to the
expansion joint premature failures experienced at PVNGS since the expansion joints
have not seen 10,000 thermal cycles. Based on past PVNGS failure data for expansion
joints, the apparent design life for the 16” diameter and 18” diameter expansion joints
are roughly 15 years. This number appears consistent with the rule-of-thumb design life
that most expansion joint manufactures use. However, the 12”, higher pressure,
expansion joints fail much faster and seem to last only 1 to 2 fuel cycles (2 to 3 years).
PVNGS has not found bellows cracking on any of the 20” diameter, lower pressure,
expansion joints.
Table 1 identifies the expansion joint sizes and configurations for the PVNGS
condenser extraction steam and drain lines. The original manufactures of the steam
expansion joints utilized single ply ASTM A240 Type 321 stainless steel for the
expansion bellows. This material was also used for the liners and covers.
Root Cause Evaluation

In 2000 a Root Cause Investigation was undertaken by a multi-discipline team on the
PVNGS January 5th 2000, Unit 2 large sodium spike event. This large sodium spike
occurred in Unit 2, 1B condenser hotwell, which increased sodium levels in the steam
generators causing a unit shutdown. Inspection of the B condenser shell on January 6th
and 7th indicated that debris from eleven extraction steam expansion joints fell onto the
1B upper tube bundle causing a condenser tube leak and subsequent sodium spike.
The expansion joint debris consisted mainly of bellows material and external bellows
shield collar rings (see Figure 1). Since the cause of the hotwell sodium spike was
easily explained by the condenser tube leak, the focus of the investigation was on the
10937790
53
equipment root cause of failure analysis (ERCFA) associated with the eleven expansion
joint failures.
The investigation results indicated that 18” diameter expansion joints 2MEDNY23 and
2MEDNY26 (see Figure 2), and their associated piping, were not designed for the
vibration service conditions in the condenser and failed due to vibration induced high
cycle fatigue. The service conditions of concern were identified as turbine vibration and
the vibration associated with internal steam flow. After the initial failure of the 18”
diameter joints, the resultant steam leakage from these joints impinged on adjacent
expansion joints causing collateral damage to nine additional joints (see RCT Photos 1,
2, 3, 4). Some noteworthy expansion joint design deficiencies were identified to include
the use of type 321 stainless steel for the bellows, the use of single ply bellows and the
lack of tie rods or hinges.
Two piping related causes were determined to have contributed to these expansion joint
failures. The first identified cause was that the expansion joints were originally not
designed for the observed vibration loads. The second was identified due to ineffective
restraint of the bellows pressure force thrust. It was recommended that the restraint of
the expansion joint center spool and piping would be required to prevent future failures.
It was also recommended that the 18" diameter vertical expansion joints be provided
with adequate restraint of the pressure thrust. Both of these issues were identified as
being addressed by replacement with hinged expansion joint assemblies (see Figures 3
& 4). The hinge design transfers the system vibration away from the bellows and into
the piping system. The bellows should then be designed so the natural frequencies are
not close to any system forcing function frequency.
One resulting corrective action from this investigation was to install a new expansion
joint design that uses hinges, has type 316 stainless steel multiple-ply bellows and has
natural frequencies that are different from the excitation frequencies. The other
corrective action was to decrease the condenser piping flexibility by adding additional
pipe guides. Corrective action effectiveness was demonstrated by performing
subsequent visual inspection preventive maintenance (PM) tasks periodically for the
expansion joints during future refueling outages.
Failure and Replacement History
PVNGS has a history of expansion joint cracked bellows that dates back to 1994 (see
Table 2). However, it wasn’t until January of 2000 that Unit 2 experienced a major
extraction steam expansion joint failure in the “B” condenser shell that required a plant
shutdown. An inspection of the “B” condenser shell indicated that 11 expansion joints
experienced major bellows damage. Bellows debris was found throughout the
condenser shell. The investigation determined that one or both of the 18” diameter
joints at the center of the condenser developed bellows failures. The subsequent steam
flow caused the remaining nine expansion joint bellows to fail due to flow induced
vibration. The investigation determined that the two 18” expansion joints were not
54 10937790
designed for the vibration service conditions in the condenser and failed due to vibration
induced high cycle fatigue.
During the subsequent Unit 3 R9 outage (fall 2001), five recently replaced extraction
steam expansion joints of an initial alternate design were found with cracks at the weld
between the bellows and the pipe spools. These expansion joints had been installed
during the previous Unit 3 R8 outage (spring 2000). Since these failures were
determined to be premature an additional root cause evaluation was performed. This
metallurgical evaluation of two samples of the fracture surfaces found that one of the
failures was due to fatigue and subsequent overload of the weld joint along the fusion
line. The joint was reported to have an unfavorable notch-like geometry where the plies
intersected perpendicularly with the weld metal. In contrast the second sample was
found to have failed due completely to fatigue. The fatigue was attributed to vibration
and/or thermal expansion and contraction.
The initial corrective action from the January 2000 Unit 2 “B” condenser event was to
replace the damaged joints with those of similar design. The follow up action was to
replace all “higher pressure” expansion joints in all three units with a newly designed
joint. The new joint is a hinged universal design (Expansion Joint Systems, Inc.) with
natural frequencies that don’t coincide with the excitation frequencies. The new
expansion joints have three ply 316 stainless steel bellows and no external shields.
Piping support changes were also implemented to stiffen the piping. These changes
have currently been implemented in Units 1 and 2. An interim design was installed in
Unit 3 and will have to be replaced with the final design in an upcoming outage.
PVNGS does not yet have in place a plan for additional periodically replacements of
expansion joints based on the empirical design life data above. PVNGS is taking credit
for the new expansion joint design and only plans to perform periodic visual inspections
and removals of these joints for internal inspection. There are no current plans to
periodically replace any other expansion joints. However, this may change based on
additional experience, evaluation, and trending.
When the root cause of failure investigation of the significant Unit 2 "B" Condenser 2000
expansion joint failure was initially completed, the only expansion joints with a history of
failures were those for the 3rd point (18” diameter) and 4th point (12” diameter) low
pressure heaters. However during subsequent refueling outages, PVNGS started to
discover cracked bellows in the expansion joints to the 2nd point (16” diameter) low
pressure feedwater heaters. Since this was not considered in the original response,
further technical and economic evaluation was developed to determine a preventative
maintenance strategy for the 16” diameter expansion joints.
Inspection and Test Methods

The original extraction steam expansion joint design included internal as well as
external bellows covers. In the early 1990’s after many of the covers had come loose or
fallen off, PVNGS decided to remove all external covers from the expansion joints inside
10937790
55
the condenser. This eliminated the potential condenser tube damage associated with
falling covers and facilitated future expansion joint inspections.
PVNGS work history indicates that expansion joint bellows inspections occurred
sporadically throughout the 1990’s. There was a detailed preventative maintenance
task developed to perform this inspection, but a set frequency was not established. It
was not clear what the basis was for performing expansion joint inspections. The
inspection consisted of a visual examination of the bellows outside surfaces using a
flashlight.
After the significant expansion joint failure occurred in the Unit 2 “B” condenser shell,
PVNGS changed its philosophy on how and when expansion joint inspections would
occur. An inspection occurred on the expansion joints inside all nine condenser shells
(three units) during every outage after the Unit 2 event. PVNGS has developed an
inspection schedule that includes a regular inspection of critical expansion joints in the
condenser neck area. It was also decided as a result of the RCF evaluation of the Unit
2 "B" event that all expansion joints with cracks or damaged bellows would be replaced.
This was a departure from the PVNGS original philosophy where cracks in bellows were
documented and allowed to remain in place for subsequent operating cycles. Based on
some qualitative assessment, these damaged expansion joints would be replaced
during a future outage.
PVNGS had proposed a Preventive Maintenance (PM) task program where all
extraction steam expansion joints are periodically visually inspected and a
representative sample of expansion joints are removed for internal inspection (see RCF
Table 3). The PM basis indicates that a visual inspection of expansion joint external
surfaces can not give evidence of damage due to fatigue, stress corrosion or general
internal corrosion. It goes on to state that where the critical nature of the system
warrants, it may be necessary to devise a means to periodically replace those
expansion joints in critical systems. However, due to the discovery of bellows cracks on
replaced expansion joints PVNGS continues to inspect the condenser expansion joints
during each refueling outage.
This strategy appears to have provided PVNGS a good indication of expansion joint
integrity the past five years and a basis on which to refine the expansion joint strategy
as needed.
56 10937790
Industry Events
INPO Operations and Maintenance Reminder, O&MR 434, was initiated in May 2001
due to a number of reports in 2000 of failed extraction steam bellows. A number of
common issues were associated with these failures:
• similar operational indications
• reduced plant efficiency or lost generation
• flow induced high-cycle fatigue contributed to failures
• bellows were approaching the end of the design life
• additional damage to adjacent components
• insufficient inspections
The extraction steam expansion joint failures described in O&MR 434 had occurred at
five nuclear power facilities including PVNGS. O&MR 434 describes similarities in
expansion joint failures between the five nuclear facilities concerning generation loss,
failure mechanisms, contributing factors, plant response and preventative maintenance
practices. The O&MR list of recommendations to consider for eliminating expansion
joint failures:
• establish and implement periodic replacement of extraction steam bellows based
on the design life of the assemblies
• use periodic visual inspections and testing to the maximum extent practicable
considering the bellows assembly design
• ensure inspections and testing procedures are specific and include acceptance
criteria
• ensure operators are knowledgeable of the operations indications of extraction
steam bellows failures
• consider removal of piping insulation inside the condenser that can become
shrapnel and damage condenser tubes
Low pressure extraction steam expansion joint failures in the condenser have continued
to occur at US Nuclear Plants since INPO O&MR 434 events were reported. A search of
the INPO Industry Events Database found four events reported in 2001, four events in
2002, one event in 2003, and two events in the first half of 2005. In addition there were
two events reported at international nuclear plants, one event in 2001, and another
event in 2002. A number of these events had resulted in down powers and repairs
during emergent or refueling outages. Many of these events have been attributed to
vibration and high cycle fatigue of expansion joint welds and bellows. Some of these
events were indicated to involve inadequate welds. Although none of these events were
identified as an impact on nuclear safety they either resulted in a reduction in power
and/or challenged the availability and reliability of power generating equipment.
Replacement and Modifications

Initially the eleven originally supplied failed low pressure extraction steam expansion
joints in the Unit 2 "B" condenser were replaced with expansion joints of an alternate
standard designs utilizing two ply 316 stainless steel bellows and tie rods with center
lugs as an interim corrective action. These interim standard designs were installed in
Unit 3 in a subsequent outage for the 12” diameter and 18” diameter expansion joints.
10937790
57
These alternate standard designs provided some additional margin to survive the
identified loads.
Since the Unit 2 Condenser "B" extraction joint failure the higher pressure expansion
joints, 12” diameter and 18” diameter, in all three Units have been replaced. The Unit 3
higher pressure expansion joints were replaced with an interim standard design utilizing
two ply type 316 stainless steel bellows and tie rods with center lugs. There have been
some case by case replacements of the Unit 3 higher pressure expansion joints with the
more robust hinged engineered design based on the results of inspections. The Unit 1
and 2 higher pressure expansion joints have been replaced with the more robust design
of appropriate natural frequencies, three ply type 316 stainless steel bellows with hinges
for the universal expansion joints.
The lower pressure extraction steam expansion joints, 20" diameter and 16" Diameter,
have been replaced on a case by case basis as a result of inspections. There are
PVNGS plans in place to replace all the 16" diameter extraction steam expansion joints.
These plans include the addition of a removable hatch with a bolted cover in each of the
three condenser shells of each PVNGS unit. These removable condenser hatches will
facilitate the work associated with future inspections and building of scaffolds and the
other activities associated with removal and replacement of expansion joints in the
future.
Operational Considerations
PVNGS Licensed Operator Continuing Training conducted simulator training on the Unit
2 expansion joint event in 2000. This simulator scenario models the failure of a 4th point
heater expansion joint failure.
PVNGS System Engineering have identified the following parameters as those

indicators experienced by Unit 2 during their expansion joint failure.
• Indications observed by Operations
• Plant Gross Electric Output Decrease
• Noise level increase in the vicinity of the main condenser shells
• LP heater shell pressure decrease
• Position of Bleeder Trip Valves (BTV)
• Additional indications observed by Engineering
• Comparison of Circulating Water (CW) temperatures across the three
condenser shells
• Condensate differential temperature across the LP heaters
• LP heater drain temperatures
Trending and Performance

In spite of the replacement of the higher pressure extraction steam expansion joints in
the PVNGS condensers the frequency of external inspections of these expansion joints
have been every 18 months. During each inspection a number of expansion joints have
been found with some cracking of the bellows and indications near the attachment
58 10937790
welds. These indications have been characterized and are replaced based on the
results.
The majority of the bellows cracking found during subsequent inspections of replaced
expansion joint has been axial. This is indicative of a hoop stress cyclic fatigue in
contrast to the circumferential cracking that is indicative of fatigue due to vibration. The
axial cracking may be attributed to pressure and thermal cyclic forces.
Additional data is being collected to determine the extent and scope of the
recommended internal inspections. These internal inspections would involve removal of
one expansion joint of each type after some operating time and then destructively
examine the joint to ensure that the joints are not developing cracks from the inside.
The results of these internal inspections would then be used to establish a Preventive
Maintenance replacement frequency with enough confidence to relax the external
inspections done every outage. These internal inspections are being considered on a
sampling basis for future outages starting in the Unit 2 R14 outage in the fall of 2006.
Benefits
The use of root cause evaluation along with industry and plant performance data to
determine the appropriate inspection and replacement techniques and intervals for the
condenser expansion joints can be an effective tool to understand failures and develop
the appropriate course of action.
Development and use of engineered extraction steam expansion joints based on current
plant data is a distinct advantage in improving plant availability and reliability. This
requires teaming with various plant disciplines, manufacturers, and consultants to
collect plant data, interpret the data, develop a design and inspection strategy, and
resolve emergent problems do to changing variables. These variables can include the
knowledge, experience, and capability of the plant staff, the manufacturers, consultants,
as well as material suppliers and handlers. In addition plant power upgrades and
changes in chemistry control, plant operations and plant maintenance can also be a
factor.
The benefits of being able to get complete life cycle of service from appropriately
designed, installed, and operated equipment is self evident. It is a careful balance of
economics and technical decision making. Study and reflection on plant problems and
sharing that information with a diverse group of plant and industry contacts can enhance
the overall knowledge of the industry as well as help validate our conclusions. This
facilitates informed decision making and is less prone to trial and error problem solving.
The value of continuous plant and equipment performance monitoring and trending can
not be overstated. This data is necessary to develop an understanding of plant and
equipment conditions and to identify system health and areas for attention.
10937790
59
Conclusions and Recommendations
The effectiveness of a multi-discipline system team made up of plant design,
maintenance, and system engineering, operations, and maintenance is valuable when
determining the most cost effective, efficient resolutions to plant problems. This team is
valuable when assessing the associated plant impacts of a course of action. These
sources help develop cost effective, budget sensitive, and timely action to plant
problems that have the potential to significantly improve plant and equipment availability
and reliability and save operating dollars.
There is great value in utilizing consultant, vendors, and industry data to help think
through and validate plant system team's thinking and conclusions at strategic points in
the problem solving process. They can bring varied information, experience, and
knowledge that help augment the plant staff. Industry data and contacts at other plants
can also be of great valve during the data collection and alternative development and
evaluation phases of the problem solving process.
It is recommended that on going monitoring and tending of condenser extraction steam

expansion joints be continued to determine the appropriate external and internal
inspection frequencies, replacement intervals, and identify those expansion joints with
indications requiring replacement prior to catastrophic failure. The continued inspection
and characterization of degraded expansion joints is required to identify additional
degradation mechanisms and the appropriate remedy. It is also recommended to
continue with plans to consider the replacement of the lower pressure condenser
extraction steam expansion joints. And finally it is recommended that plant personnel
stay current on the developments in condenser and expansion joint technology and
plant heat rate improvements to more effectively address plant secondary issues as
they emerge.
References
1. EPRI Final Report 1008035 (May 2003), Expansion Joint Maintenance Guide
Revision 1
2. EPRI Final Report 1003288 (December 2002), Repair of Metal Bellows
3. Expansion Joint Manufactures Association, Inc., Standards of the Expansion

Joint Manufactures Association (1998)
4. INPO Operations and Maintenance Reminder (O&MR 434) May 21, 2001,
Extraction Steam Bellows
5. INPO Industry Events Database
6. PVNGS Maintenance Engineering White Papers (2001- 2003), Primary Author

Tim Szumski (unpublished)
60 10937790
Expansion Expansion Tie Rods LP Extraction Operating Operating
Joint Joint Type Heater Stage Temperature Pressure
Diameter Point (oF) (psia)
20" Vertical No 1 12 195.5 10.5
Universal
16" Pressure Yes 1 12 195.5 10.5
Balanced
(Drain Line)
16" Vertical Varies 2 11 232.0 21.5
Universal
16" Horizontal Yes 2 11 232.0 21.5
Single
Bellows
12" Pressure Yes 2 11 232.0 21.5
Balanced
(Drain Line)
18" Vertical No 3 9 329.0 68.7
Universal
28" Pressure Yes 3 9 329.0 68.7
Balanced
12" Vertical Yes 4 8 419.0 114.1
Universal
18" Horizontal Yes 4 8 419.0 114.1
Universal
18" Pressure Yes 4 8 419.0 114.1
Balanced
Table 1: Condenser Expansion Joint Configurations
10937790
61
End Plate
Tie Rod
External Bellows Shield
Collar Ring
Center Lugs Center Pipe
Inner Bellows Bellows

Shield
Flow
Figure 1
Expansion Joint Sketch
62 10937790
North
Y17 Y18 Y19 Y20
Minor Steam Wave
Y27 Y24 Y26 Y28
Y67
Y23 Y25
Y22 Y21
Major Steam Wave
Y32 Y31 Y30 Y29
Key
- Completely Damaged Expansion Joint

Expansion Joint Sizes (Diameter)
Size Equipment
- Badly Damaged Expansion Joints
12” Y24, Y25
16” Y21, Y22, Y27, Y28, Y67 - Damaged Expansion Joints
18” Y23, Y26
20” Y17, Y18, Y19, Y20, Y29, - Undamaged Expansion Joints
Y30, Y31, Y32
- Y17 means Expansion Joint 2MEDNY17 (typical)
Figure 2
Plan View of Expansion Joint Layout in 2B Condenser
10937790
63
RCF Photo 1
Northeast Side of Expansion Joints
2MEDNY23 (right) and 2MEDNY26 (left)
RCF Photo 2
Southwest Side of Expansion Joints
2MEDNY23 (left) and 2MEDNY26 (right)
64 10937790
RCF Photo 3
View of Expansion Joints Looking Northwest
RCF Photo 4
View of Expansion Joints Looking Southwest
10937790
65
Hinge Ring Hinge Pin
Hinge
Attachment
Pipe
Figure 3
Hinged Expansion Joint Design
66 10937790
Figure 4
EJS Hinged Expansion Joint Design (Used with permission)
10937790
67
EQ ID Size 2R4 S93 2R5 S94 2R6 S96 2R7 F97 2R8 S99 SNOW
Work Inspection Inspection Inspection Inspection S2000
EDN-Y17 20” Removed covers Replaced
EDN-Y18 20” Removed covers
EDN-Y24 12” Replaced Cracks 6 cracks Replaced
EDN-Y25 12” Removed covers Cracks Cracks 8 cracks Replaced
EDN-Y56 12”
EDN-Y59 16”
EDN-Y71 18” Cracks
EDN-Y74 28” Cracks
EDN-Y50 8” 11/90 Failed Joint NA NA
11/91 Replaced eliminated
EDN-Y53 10” Joint NA NA
eliminated
Table 2: Unit 2 B Condenser Expansion Joints
68 10937790
Unit 2 Unit 1 Unit 3
Outage
Shell A Shell B Shell C Shell A Shell B Shell C Shell A Shell B Shell C
9 Install Partial Install Install Install Install
10 Finish Install Install Install
11 Visual
12 ** Visual
13 Visual
14
15 Removal
16 Visual
17 Visual
18
19 Visual
20 Removal
21 Visual
22
23 Visual
24 Visual
25 Removal
Table 3: Expansion Joint Preventive Maintenance (PM) Strategy
Key
Install - Installation of EJS expansion joints

Partial Install - Partial Installation of EJS expansion joints
Finish Install - Finish Installation of EJS expansion joints
Visual - Visual Inspection of all expansion joints
Removal - Remove and replace each type of EJS expansion joint (6 types) with a new or rebuilt
joint.
Destructively examine all removed expansion joint bellows looking for cracks.
** - Finishing up the installation in the Unit 2 “B” shell can occur anytime during this time frame.
10937790
69
10937790
2
SESSION 2: PERFORMANCE TECHNOLOGY
10937790
71
10937790
EPRI Condenser Technology
Seminar & Conference
8/30-9/1, 2005
San Diego, CA
On a Predictive Condenser Performance Model
Joseph W. Harpster, Ph.D.

President, Intek, Inc.
751 Intek Way
Westerville, OH 43082
T: 614-895-0301, F: 614-895-0319
E: jharpster@intekflow.com
ABSTRACT
A review of historical condenser design methodology is presented from the perspective of using
average tube performance behavior. Conditions are described showing that assumed average values
should not be employed to quantify measured performance indicators, since the cause of performance
degradation can be masked. Plant measurement data will be provided to demonstrate how
performance computations can have significant error.
A new predictive condenser theory and analytical model, having patented and patent pending
methods and condenser design features, is described that anticipate causes for condenser
performance losses and how they may be circumvented. It is shown that this approach does not
depend on special operating condenser measurement results to establish boundary conditions and
condensation rate patterns as has been the case for the currently employed Computational Fluid
Dynamics (CFD) method. This new approach permits the application of computer modeling,
employing the new comprehensive and predictive model and theory, along with normal performance
data, to reproduce, as close as possible, measured performance results. Pleasing portrayals of
velocity profiles generated by Finite Element Analysis (FEA) is presented showing the fineness of
this new, predictive, method and its utility.
INTRODUCTION
Condensers have seldom achieve the operating or performance specifications required of them
during design phase. It has generally been accepted that during the period of fabrication,
construction and startup the condenser’s tube heat transfer coefficient can be compromised due to
contamination of the tubes. This heat flow barrier is believed to be mostly caused by deposits and
biofouling on the cooling water side from the initial exposures to circulating water contaminants.
An easily recognized performance indicator responsive to the tube fouling is the measured condenser
pressure, which would be in excess of the design pressure under a fixed set of operating conditions.
For a particular condenser, the primary operating parameters that can affect the pressure P are the
10937790
73
cooling water velocity in each tube vcw, temperature of the inlet cooling water Tcw, quantity of steam
to be condensed Q, air in-leakage into the shell space Va, and cleanliness of the tubes. With all of
these known at a given time, including insignificant amounts of air in-leakage, and known clean
tubes, an appropriate design pressure PD can be determined using the HEI method. This method
employs application of the tube heat transfer coefficient UHEI which generally includes discounting
to a design value UD, by employing a cleanliness factor Fc. This factor is presumed to have no
attributes related to fouling(1) since it only applies to a clean condenser. HEI does state however, that
circulating water and condensing steam will produce heat transfer resisting films on tube surfaces
that will have characteristics related to both the type of fluid and the material(2) . . . these should be
considered in selecting a suitable cleanliness factor as a design coefficient. Should P be greater than
the predicted PD, even though an Fc has been employed, then it has been sometimes concluded that
the cause is from the contaminating cooling water, and that the tubes are indeed unclean.
The design cleanliness factor applied to UHEI should reflect the operating heat transfer value that is
experienced in practice for a new, clean condenser. Because of this, Tsou(3) recommended that the
term Performance Factor PFdes be used in place of Fc. Earlier, in a publication by Marto and Nunn(4),
addressing the dilemma facing condenser designers and researchers attempting to develop an
adequate theory to describe the shell side steam/air flow behavior, McNair(4) in the preface referred
to these correction factors as "factors of ignorance," explaining that they lead to (unexplained) over-
design of condensers which were considered a great liability in size and weight, particularly for
ocean going ships, both merchant and Naval. For this reason, justification was implied for
conducting further research to understand the unknown cause or causes requiring that this important
coefficient be derated.
A large number of operating condensers having clean tubes and insignificant amounts of air in-
leakage were tested. The results were that design pressure was more than likely not achieved, and
measured Heat Transfer Coefficient Umeas, fell below the design value UD, even when tubes were
clean. The implication, recognized as far back as 1960 by Gray(5) and which continued for the
remainder of the 20th century, is that the design practice employed is inadequate. Little progress has
been made since the early 1980s to improve predictive condenser modeling and this includes the
current efforts using Computational Fluid Dynamics (CFD) modeling. CFD has been used with
some success to simulate measured outlet water box thermal stratification caused primarily by air
bound regions in many older condenser tube bundles. Usually however, this modeling would be
performed after actual temperature rise measurements were made for an array of tubes in the outlet
water box, thus establishing a sort of boundary condition. Therefore, this practice has placed the
predictive capability of CFD in question.
With the introduction of a condenser diagnostic instrument(6,7) hereafter referred to as a MultiSensor

Probe (MSP), new information on the water vapor/noncondensable gas mixture being extracted from
the condenser by the vacuum pump was measured. This provided the author an opportunity and the
ability, for the first time, to examine the causes that affect the composition of this mixture originating
deep within the condenser tube bundle. The result was a model describing condenser behavior and
a comprehensive theory describing the dynamics(8) for pure steam and steam containing non-
condensable gases. Further work has identified additional features of noncondensables gases, even
at low air ingress, responsible for poor condenser performance and affecting the determination of
Fc,meas and PFmeas values, the respective measured values of cleanliness and performance factors.
74 10937790
This paper reviews the author’s comprehensive condenser model showing that previously
unrecognized condenser design issues can be identified as appropriately responsible for the noted
pressure discrepancy. Also discussed is how the measurements needed by the HEI method for
determining the effective heat transfer coefficient Ueff, at its current operating condition, can have
error caused by the design problems. Excess pressure and other deleterious consequences of
condenser performance are traceable to these design deficiencies.
CONDENSER MODEL OUTLINE
Since the first application of the MSP in 1994, the author has been examining new measurement data
previously unknown to other investigators.(9) Correlation of data from many plants and a wide
variety of condenser configurations has led to an understanding of commonly observed behaviors
in steam-air dynamic phenomena, despite the wide differences in condenser construction. The
results have been used over the past several years to support condenser diagnostics to understand the
cause of performance related issues. These include the interrelated condenser performance indicators
such as high pressure, low cleanliness factors, low heat transfer coefficient, and high Dissolved
Oxygen (DO); along with their contributions to high heat rate, excessive forced outages, and high
iron transport.
The theoretical model is based on six (6) simple logical behaviors that are used to illuminate what
has been perceived as the very complex nature of shell side steam dynamics.(8,9) For almost all
condensers in use today, these behaviors are:
1. Steam, without air, entering the tube bundle from all sides perpendicular to the tube’s length
in an approximate velocity range of 20 to 40 feet per second, is condensed on tubes at a rate
that remains essentially constant up to the last tubes near the center of the bundle, where the
velocity may become zero.
2. Any air entering the condenser shell space, through a leak from outside the shell, gets caught
up in the steam, is scavenged into and through the bundle, and is transported toward the tube
bundle central region. The consequence is that its concentration is increased along the path
length, which, at high concentration, interferes with the condensation rate.
3. If the leak is small, the air may be transported to an air removal section and is removed by
the condenser exhausters, with essentially no consequence to condenser back pressure and
very little dissolved gases entering the hotwell condensate.
4. If the leak is large, the concentration of air near the center of the bundle becomes sufficient
enough that a significant partial pressure of air will develop at that location. For this to
happen the affected tubes in this space must drop in temperature. It does so by reducing the
rate of steam condensation on these tubes, allowing the condensate on tubes to be further
cooled by circulating water in the tubes, thus establishing a reduced water vapor pressure
below that of the steam. The pressure in the region becomes essentially the sum of the partial
pressure of air and the partial pressure of the lower temperature water vapor. This low
temperature region with high partial pressure of air is a major source of DO in condensate.
This total pressure is, for all practical purposes, in equilibrium with the steam pressure
outside this region, and steam is severely restricted from entering therein. The region is
termed the stagnant zone, because of the lack of “dynamic” steam flow.
5. The effect of the nearly developed and fully developed stagnant zone is to reduce the number
of active tubes in the condenser. This has the consequence of raising the condenser pressure.
10937790
75
6. The pressure rises until the exhauster suction pressure at the stagnant zone becomes
sufficient to permit its capacity for removing air from the condenser to match the amount of
air in-leakage, where the pressure then remains stable.
The above is a natural and easily understandable consequence of steam and air mixture dynamics
resulting from air in-leakage into a condenser. However, a similar phenomenon, identified as Air
Binding (AB),(10) will be discussed in the following section. AB is a direct result of condenser design
deficiencies that are correctable,(11) even in the absence of significant air ingress, and is also a source
of DO. Patented and patent pending methods(12) related to condenser modifications and
improvements in new design, resulting from the above and other behaviors are now known and can
be used to minimize or eliminate the effects of these deficiencies. These methods are developed
based on the new theoretical model alone and are not the consequence of condenser modeling using
computers employing tools like CFD, FEA or other engineering or scientific software programs.
AIR BINDING EFFECTS
The detailed cause of AB and its consequences were reported earlier by Harpster.(10) AB has been
known to happen when air or other noncondensable gases are collected in pockets within sections
of the tube bundle. However, it has only recently been understood from the comprehensive model(10)
that these pockets result predominantly from the scavenging process as described above, but in the
case of AB, the collected air cannot be directly removed by the air removal section. Figure 1
provides an example of air binding in an operating condenser showing its effect on measured non-
uniformity of temperature rise in the cooling water passing through an array of tubes in a condenser.
This data leads to the obvious conclusion that the heat transfer coefficient in the regions outlined is
diminished. The temperature rise of tube circulating water in this region is small, because there is
a reduced amount of steam being condensed on these tubes. The heat transferred to the tubes is
related to the steam temperature by the well known Fourier equation, given by,
Q = U A ∆ Tlm (1)
where ∆Tlm is the logarithmic mean temperature difference, which has a dependence on the saturated
steam temperature, on the circulation water temperature inlet and outlet values. Q is the heat input
per tube having a contact surface area A and a heat transfer coefficient U, the latter remaining
constant for non-air-bound tubes but will vary inside the AB zone. The rise in steam temperature
resulting from AB zones is similar to that of the stagnant zone and is accompanied by a rise in the
condenser pressure, reducing generator capacity, or requiring more fuel to the boiler to maintain
load.
In addition to increasing condenser back pressure, an AB region

can also affect condensate water chemistry. The reduced
condensation rate in the AB zone will reduce the tube
temperature rise along the length of tubes in the zone. This will
result in temperature differences (referred to as condensate
subcooling) between the surface temperature of water on the
tube and the condenser steam, or the steam in the gap between
tubes outside the AB zone. Some of this space between the
tubes and near the tubes in the AB zone will contain water vapor Figure 1
Temperature rise for one of eight
identical bundle sets
76 10937790
having a saturation temperature and
vapor pressure below that of the 600
surrounding steam. This in turn permits PO2 = 0 .2 Pnon condens ables
500
Dissolved Oxygen, PPB

air or other noncondensables to exist,
having their own partial pressure 400
consistent with the heat transfer 300
modifications obtained by Henderson 200

and Marchello.(7) The AB region is
100
dynamic and will contain thermal layers
of steam and various amounts of water 0
0 5 10 15 20 25 30 35 40
vapor and air, in varying ratios, unlike Stagnant Zone Subcooling, °F
the stagnant zone. This condition
explains the existence of AB regions Figure 2 - Upper Limit of DO vs Subcooling
using the new theory(2,4) and is not the in Condenser Stagnant Zones at 85°F
condition applicable to the rest of the Inlet Cooling Water Temperature
condenser, as was thought by many
during the 20th century.
Since there exists a partial pressure of gas near tubes in the AB zone and the surface temperature of
local condensate in the AB zone of tubes is subcooled, like that in stagnant zones, these gases can
be absorbed into this water approaching a concentration X, dictated by Henry’s law governing
solubility(8), given by,
Xi = Hi pi (2)
where Hi is Henry’s law constant, which has a temperature dependency, and pi is the gas partial
pressure.
Therefore, associated with AB, there will be dissolved gases in this local condensate. The partial
pressure can be expressed in terms of the amount of subcooling. Figure 2 presents the equilibrium
concentrations of oxygen based on the amount of subcooling. Thus, not only excess back pressure
but dissolved corrosive gases as well can result from condenser designs that promote AB zone
development.
Under conditions of clean tubes and low air in-leakage, it would appear that the experience-driven
use of a design “Cleanliness Factor” Fc resulted from the inadvertent existence of AB zones in a
condenser. It should therefore be clear that the use of a discounted heat transfer coefficient may have
been required to overcome the effects of AB, and perhaps other unknown conditions, in order to
meet design requirements. Unfortunately, for many condensers, the measured heat transfer
coefficient Ueff under conditions of low air in-leakage and clean tubes still fall below the values of
UD, the discounted HEI heat transfer coefficient UHEI, having the relationship; UD = Fc UHEI
EFFECTS OF AB ZONES ON MEASUREMENTS
Design conditions causing AB zones have a significant effect on condenser performance, and on
performance indicator measurements. The accurate measurement of operating parameters that are
necessary to compute additional performance indicators, such as the effective Heat Transfer
10937790
77
Coefficient Ueff, are also compromised by the presence of AB zones. Measured values of parameters
necessary to compute performance indicators employing averaging methods, either implied or stated,
using the HEI method, may yield gross error in the determined indicator.
Figure 1 shows an operating condenser tube bundle containing an AB zone and the resulting
measured circulating water temperature rise array pattern. The temperature rise of the lower middle
region is much smaller than the high value shown for the upper and side regions, which have an
average value of 24°F. This lower temperature rise region is an AB zone where the heat transfer
coefficient is steadily diminished to near zero due to the presence of entrapped air, or
noncondensables. This condition leads to the industry-wide recognized, but sometimes
misunderstood, outlet water box temperature stratification. Many operating plants attempt an
accurate determination of the average outlet water temperature by using multiple temperature sensors
in the outlet water box.
Should the average value of two temperature sensors located in the outlet water box be used, it is
easy to understand that a range of erroneous values for a determined circulating water temperature
rise ∆Tcw could be obtained. Table 1 shows some selected possibilities based on temperature sensor
locations as shown for the condenser presented in Figure 1. The last column shows the magnitude
of error if the measured average of these two sensors that may be picked and then compared to the
real weighted average value for ∆Tcw, which was 18.4°F for the condenser section shown in Figure
1.
Table 1 — ∆ Tcw Determination for Tube Bundle of Figure 1

with different locations of two temperature sensors
(Real weighted average ∆Tcw = 18.4°F)
Selection of
two locations
∆ T1 ∆T2 Average ∆Tcw Error %
1 24 0.5 12.25 -33.4
2 19 14 16.50 -10.3
3 24 7 15.50 -15.8
4 19 7 13.00 -29.3
5 24 19 21.5 16.8
6 24 24 24 30.4
7 19 19 19 3.3
8 19 .5 9.75 -47.0
9 24 14 19 3.3
In computing Ueff and a performance cleanliness factor Fc,eff, a value for ∆Tcw must be used. If the
method used requires the thermal load on the condenser to be determined using this temperature
differential, then the magnitude of error in this critical computation would be directly proportional
to the error found in the final column of Table 1. However, in subsequent computations of TTD, and
78 10937790
finally Ueff, the percent error increases to where it can be more than doubled for the determined value
of Ueff.
The above illustrated error in Ueff and consequently to Fc,eff is all due to condenser design issues
related to AB. Obviously, tube fouling and air in-leakage would further modify the determined value
for Ueff and these could affect the error distribution shown in Table 1 for the various sets of sensor
location provided.
CORRECTION METHODS FOR AB ZONES
The cause of AB zones can be significantly reduced or eliminated using patented or patent pending
techniques and methods.(12) These employ the use of strategically located baffles, piping and
structures to take advantage of natural steam/air mixture dynamics in the condenser that affect the
AB zones and air removal sections. These described techniques and methods are independent of any
particular analytical approach used to describe steam and air flow patterns around and through the
tube bundle. The approach does not make use of removing tubes from the outer perimeter of the
bundle to create steam lanes.
The uniqueness of the approach taken is the employment of the comprehensive condenser theory and
model(8,9) permitting the establishment of operating boundary conditions for use with computer
modeling programs. The theory alone is sufficient to identify locations of AB zones, allowing fixes
to be estimated. The modeling then defines precise location(s) and extent of modifications.
The result of this effort is not only to minimize condenser pressure through improvement of Ueff but
also to reduce dissolved oxygen, reduce corrosion and forced outages and to extend plant life. Other
related benefits are reduction in CO2 emissions, lower heat rate, reduced fuel usage, improved
capacity for base loaded units, and reduce or obviate load reductions during summer time operations.
Following condenser modifications a new set of drawings identifying the changes for future
reference are provided along with a new set of performance data sheets, a nameplate to be attached
to the structure identifying the modification and a license to use applicable patented methods on the
identified condenser. An average effort to implement changes during an outage is estimated to be
about ten (10) man-days.
MODELING
When it is necessary or beneficial to model steam/air flow profiles throughout the condenser,
boundary conditions for this purpose are established by making use of plant performance data and
application of the author’s comprehensive condenser model. Although air in-leakage can be
variable, it is generally set at the maximum level in units of SCFM appropriate to the threshold value
necessary for the onset of condenser excess pressure. This modeling includes performance within
the air removal section and is not limited to a fixed water vapor to air mass ratio entering the air
offtake vent line believed by some as being prescribed by HEI.
Individual slices of the condenser perpendicular to the tube length are modeled using the above well
defined boundary conditions. The modeling is performed using FEA methods.
10937790
79
Figure 3 shows the steam flux density through three sections
of a complex single pass tube bundle. The effects of AB is
obvious and increasingly worse in the direction of cooling
water flow with little steam flow throughout most of the
outlet end section. These results are derived more from
accurate determination of boundary conditions than from the
artistic picture drawing capability of the modeling program.
CONDENSER MODIFICATION PROGRAM
Experience shows that a very high percentage of condensers

suffer from design deficiencies. Studies have been
conducted on several condensers where excess pressure, due
to inadequate condenser design or construction, was found
to be in the range of 0.2 to 1.0" HgA. This corresponds to
heat rate excess of 0.34% to 1.7%
Design issues identified from the model and condenser

inspections are AB zones and inadequacy of the air
removal section of the condenser. A program to identify
and then to minimize effects of these problems has been
developed. This program has a minimum of four major but
independent study phases, conducted sequentially.
Initiating work on a subsequent phase may depend on the Figure 3.FEA Modeling showing
results of the previous phase. Particular attention is given steam/air flux in air bound condenser
to cost benefit evaluation.
Program Overview
Phase 1: Condenser drawing review to assess the possible existence of AB and adequacy of
the air removal section. Condenser inspection may be necessary.
Phase II Engineering evaluation to determine the extent of the problem, including condenser
inspection, and tests.
Phase III Modification engineering to define the extent of needed changes or modifications that
can be made to the condenser to eliminate or significantly reduce the effects of AB.
Phase IV Follow up to quantify the improvement through measurements.
The program’s modifications can be typically executed in connection with normal scheduled outages.
SUMMARY
Historically, condenser manufacturers employed experience based performance results as a guide

for modifying condenser design with the hope of capturing performance improvement. This paper
summarized a new 21st century comprehensive theory and model that explains all performance and
performance limiting features known to exist in operating condensers. This new theory also
predicted behaviors that were later identified by observation on operating condensers. It was noted
that this model and plant performance measurements are sufficient to establish boundary conditions
80 10937790
necessary for accurate description of steam/air mixture mass flux densities and flow patterns using
computer aided modeling.
The model and theory alone, without the necessity of computer aided modeling, was shown to
predict AB zones resulting from condenser design deficiencies. The strength of modeling is to
identify sharply the dimensional characteristics of modifications and to show pictorially the steam/air
mixture flux density incorporating a direction of flow.
It was shown that AB zones along with air in-leakage are causes for thermal stratification in the
outlet water box that can cause great error in performance calculations due to a lack in defining a real
value for the average temperature rise of circulating water. This error is significantly increased when
target performance indicators such as heat transfer coefficient and cleanliness factors are calculated
from these measurements.
A program was presented to identify specific modifications unique to individual condensers using
patented methods. Any modifications made will be documented in a new drawing set and a separate
nameplate provided for attachment to the condenser indicating the date of modification. New
performance data sheets are provided along with description and drawings of the modification. So
far, modification performance improvement with respect to back pressure is in the range of 0.2 to
1" HgA. The annual economic value associated with this alone can justify costs of the study and
modification. Many other even higher value benefits are a direct result of these modifications
through improved condensate cycle chemistry. Outages and chemical cleanings are reduced; and
plant life is extended.
REFERENCES
1. Putman, Richard E., Steam Surface Condensers, ASME, NY, 2001.

2. HEI, Standards for Steam Surface Condensers, 8th Edition HEI Cleveland, OH 1984.
3. Tsou, John, "New Methods for Analyzing Condenser Performance." Proc. 1994 EPRI Heat Rate Improvement
Conference, Baltimore, MD, May 1994.
4. Marto, P.J. and Nunn, R.H., Power Condenser Heat Transfer Technology, Hemisphere Publishing Corporation,
Washington, 1981.
5. Gray, J.L., Discussion pp. 358-359, on Silver, R.S., "Approach to a General Theory of Surface Condensers." Proc.
of the Institute of Engineers, Vol. 178 Part I, London, England. 1963-1964.
6. Harpster, Joseph; Molnar, David and Thompson, Dean "Advancements in Air In-leak Measurement." Power-Gen
94 Proceedings, Orlando, Fl., Dec. 7-9, 1994.
7. Putman, Richard, Condenser In-Leakage Guideline, EPRI, TR-112819, January 2000.
8. Harpster, Joseph, "An Impact on Plant Performance from Advanced Instrumentation." ISA POWID, Orlando, FL.
July 7-12, 2001.
9. Harpster, J.W., "On Understanding the Behavior of Noncondensables in the Shell Side of Steam Surface
Condensers." ASME International Joint Power Generation Conference, New Orleans, LA, June 4-7, 2001.
10. Harpster, J.W., "Increased Availability from Improved Condenser Design." ICONE-10 International Conference
on Nuclear Engineering, Arlington, VA, April 14-18, 2002.
11. Harpster, J.W., "Reducing Dissolved Oxygen under Conditions of High Air Ingress." ASME International Joint
Power Generation Conference, Phoenix, AZ, June 24-26, 2002.
12. Harpster, J.W., USPN 6,526,755 and pending patents.
10937790
81
10937790
A Study of the Effect of Plant Variables on the Performance the Pilgrim Condenser
Norman Rhodes and Charles D Hardy
Heat Exchanger Systems, Inc.

86 Finnell Drive
Weymouth, MA 02188
Abstract
Entergy’s Pilgrim condenser has two shells connected with a steam transfer duct. Within each shell there
are two pairs of tubesheets, each pair of tubesheets contains two tube bundles. Although the two shells
are identical in design the pressure in each shell and the cooling water temperature rises measured in each
bundle have been found to vary. For example, mean temperature rises in one shell have been in the range
33 to 35F while the other shell is about 25 to 26F.
This paper describes a study carried out to assess what plant operational factors might cause this variation
in cooling water temperatures. A detailed three-dimensional model, based on Computational Fluid
Dynamics (CFD) techniques, has been developed and applied to predict the performance of the
condenser. The model, being based on fundamental principles and a highly detailed resolution of the
condenser, enabled different plant conditions to be critically assessed to determine whether they might
lead to these situations. The model was developed and validated by application to known steady summer
and winter plant conditions provided by Entergy. Following this stage, the model was applied to study the
effect of a wide range of inlet cooling water temperatures, and for the following plant variables:
• Impact of inoperable water box air scavenging system which results in the outlet water box being
partially filled to 18” below top of tube sheet.
• Impact from moisture separator(MS) Dump valve Steam Leakage of 1%, 2.5% and 5% at 200
PSIA entry in West Condenser
• Impact from tube plugging of 2, 5, 10, 20 and 30%.
• Impact from air in-leakage from 20 to 50 scfm.
1 INTRODUCTION
This paper describes a study to predict the flow and heat transfer in the Pilgrim condenser. A three-
dimensional model, based on Computational Fluid Dynamics (CFD) techniques has been developed and
applied to predict the performance of the condenser.
The general layout of the Pilgrim condenser is shown in Figure 1. Two CW pumps, A and B, supply to
one waterbox in each shell. The four waterboxes are denoted 1-1, 1-2, 1-3 and 1-4. Hence, CW pump A
supplies to 1-2 and 1-4, and CW pump B to 1-1 and 1-3.
Although the two shells are identical in design the pressure in each shell and the CW temperature rises
measured in each bundle are known to vary according to data provided by Entergy. Figure 2 shows one
such variation which occurred during December 2003. The higher frequency oscillations in the results are
due to tidal effects on the CW system flow rates. It can be seen that the mean temperature rises were
initially in the range 25 to 30F. Over the course of about two days a divergence occurred during which
temperature rises in the East Shell rose to values of about 33 to 35F while the West Shell values fell to
about 25 to 26F.
10937790
83
Figure 1: General layout of condenser shells and CW arrangement
Figure 2: Measured Temperature Rises in the East and West Shells
The study was carried out to determine the effect of different plant conditions that might lead to these
situations. A CFD model was developed and validated by application to known steady summer and winter
84 10937790
plant conditions provided by Entergy and then used to assess the effect of possible plant variables,
specifically:
• Impact of inoperable water box scavenging system which results in the outlet water box being
partially filled to 18” below top of tube sheet.
• Impact from MS Dump valve Steam Leakage of 1%, 2.5% and 5% at 200 PSIA entry in West
Condenser
• Impact from tube plugging of 2, 5, 10, 20 and 30%.
• Impact from air in-leakage from 20 to 50 scfm.
2 PHYSICAL SITUATION
The Pilgrim condenser comprises two shells and each shell contains two identical tube bundles. The
condenser has a single pass arrangement. Figures 3 and 4 illustrate the main internal details of one shell.
Figure 3: Cross Sectional elevation of one condenser shell
Figure 3 shows a cross-sectional view of the west condenser shell, viewed from the CW inlet end, and
illustrates the two feed-water heaters in the upper part of the shell, the support plates and the outline of the
tube bundles. Each waterbox (1-1, 1-2 etc.) contains one tubesheet with two symmetrical bundles, hence
there are four tube-bundle sections in each shell. An air offtake point is located the CW inlet end of each
bundle.
10937790
85
The main dimensions and properties of the condenser are as shown in Table 1.
Parameter Value
Number of Tubes 17544 per shell
Effective length 49.8 ft
Tube diameter 7/8 in
Tube thickness 22 BWG Ti ASTM B388 GR2
Tube Pitch 1.125 in
Thermal Conductivity 12.68 Btu/hr·ft²·°F/ft
Table 1: Parameter used in the model
3 CONDENSER MODEL
The fundamental equations which are used to describe fluid motions are based on the laws of
conservation of mass, momentum and energy. The equations are solved using a finite-volume method
which involves the subdivision of the flow domain into small volumes, or cells, and the derivation of
algebraic finite-domain equations by integration of the above differential equations over each grid cell. The
resulting equations are then solved iteratively, for prescribed boundary conditions, thus obtaining values of
the flow variables in each grid cell and hence throughout the condenser.
The following transport equations are solved in the model:
• Steam/air mixture velocities in three co-ordinate directions (u, v and w)
• Pressure, p
• A scalar representing the steam/air mixture c.

The ratio of steam to non-condensable gas concentration, which is hereafter referred to simply as the
steam/air concentration is defined as:
msteam
c=
msteam + mair
The computational model for the flow of steam through the condenser includes heat and mass transfer, the
effect of inundation (the reduced heat transfer due to the increase in water film thickness over the tubes
caused by condensate falling from tubes higher in the nest) and frictional pressure drop on all surfaces,
such as baffles and support plates, and due to the tube nest. These parameters are calculated in each grid
cell using the locally predicted variables. Thus, in contrast with the normal design process where a
constant heat transfer coefficient is assumed, the heat transfer coefficient varies throughout the condenser
and is influenced by the local conditions. The condensation and friction processes are represented as a
series of momentum, mass and scalar sink terms in the appropriate equations as described in Reference 2.
Figures 4 shows a perspective view of the condenser model outline. The East and West Shells, the
interconnecting transfer duct, the tube support plates and the eight air offtake ducts which run along the
length of the condenser at the top of each bundle, can be seen.
86 10937790
Figure 4: Perspective view of Condenser Model
In each shell a grid of 82 x 49 cells has been used in the x- and y-coordinate directions respectively (the
cross section represents an x-y plane in the model with the y-coordinate being the vertical coordinate). In
the z-coordinate direction 42 cells are used. The cooling water inlet end is at z=0. In the transfer duct a
grid of 10 x 7 x 9 has been used. Thus the model utilizes 338,142 cells to represent the condenser
geometry.
The cells which contain tubes are tagged with an identifier which describes the orientation of the pitch
and the presence of lanes where applicable. In the process of the calculation these identifiers are used to
evaluate the total surface area for heat transfer and the number of tube rows in the cell to adjust the
pressure drop terms.
3.2 Boundary Conditions
To complete the model, the following boundary conditions were applied:

• Turbine Exhaust: Steam/Air inflow from the turbine is defined as a mass flow rate of steam and a
steam/air concentration at the condenser inlet plane. In the model, the inlet velocity is prescribed as
the boundary condition and is calculated from the prescribed mass flow and the predicted condenser
pressure. Hence, the magnitude of velocity is updated as the solution proceeds.
• Cooling Water Temperature and Flow Rate: The cooling water inlet temperature and mass flow
rate are set at the specified conditions.
• Air Offtake: The air offtake is defined as a constant pressure boundary. It is important to note that
the air removal equipment is not modeled explicitly.
4 PRESENTATION OF RESULTS
4.1 Typical Model Results
The three-dimensional model provides a detailed representation of the flow and heat transfer within the
condenser. This section presents typical results obtained with the model in the form of flow-visualization
10937790
87
plots which show the general flow behavior. The following results are taken from a base-case run
performed with an inlet CW temperature of 35F .
Figure 6 shows the velocity of the steam/air mixture at a mid-plane along the length of the West Shell.
The steam accelerates from the inlet, around the feed-water heaters, and high flows are observed between
the tube bundles. Within the tube bundles the velocities are relatively low. The steam transfer duct is
shown in this plane; for this case the transfer flow is small.
Figure 6: Velocity Vectors in West Shell mid-section
Figure 7 and 8 show pressure and steam/air concentrations in both shells at the mid plane. Highest
pressures occur when the steam approaches the feedwater heaters and the top of the tube bundles.
Pressure within the tube bundles falls progressively toward the air offtake duct, where the pressure is
lowest.
The steam/air concentration plots, Figure 8, shows that the steam is generally at high concentrations
except in the air offtake duct.
88 10937790
Figure 7: Pressure Contours at mid-section of condenser
Figure 8: Steam/Air Concentration Contours at mid-section of condenser
4.2 Model Verification
Two cases were chosen for verification of the model, based on plant operating data provided by
Entergy which was taken on 11/26/03, a winter condition, and 8/18/03, a summer condition. The
boundary conditions which have been applied are as follows:
CW inlet temperature-winter 48.2oF* (11/26/03)

CW inlet temperature-summer 67.1oF* (8/18/03)
Total CW flow rate (both shells) 311,000 gpm
East Shell CW flow 49.7%
West Shell CW flow 50.3%
10937790
89
LP turbine inlet mass flow 4,878,135 lb/hr
Steam inlet enthalpy 963.4 BTU/lb
Fouling factor 0.85
* Temperatures are the average of the “A” and “B” CW pump inlets
The flow difference between the East and West Shells is based on information provided by
Entergy and is due to the inlet/outlet flow conduit configuration that feeds in from the West side
and runs across to end at the East Shell water boxes.
The measured plant conditions were as given in Table 3:
East West 1-1 outlet 1-2 outlet 1-3 outlet 1-4 outlet
Turbine Turbine CW temp CW temp CW temp CW temp
Back press Back press rise (F) rise (F) rise (F) rise (F)
(inHg) (inHg)
Winter 1.74 2.04 25.4 26.3 29.1 26.5
Summer 2.6 2.65 23.8 24.6 27.3 25.6
Table 3: Measured Plant Conditions
It was suggested by Entergy that the benchmark cases should only utilize condenser pressure
data only from the East Shell, based on the conclusion that the West Shell pressure data is
inaccurate.
The values predicted by the model are as follows:
East West 1-1 outlet 1-2 outlet 1-3 outlet 1-4 outlet
Turbine Turbine CW temp CW temp CW temp CW temp
Back press Back press rise (F) rise (F) rise (F) rise (F)
(inHg) (inHg)
Winter 1.68 1.68 27.6 27.8 28.6 27.3
Summer 2.62 2.62 27.8 27.7 28.1 27.7
Table 4: Predicted Values for Summer and Winter Cases
It can be seen that the pressures predicted by the model are close to the measured values for the
East shell. There is no difference in pressure between the shells, as would be expected given the
slight difference in CW flow rates. The predicted CW temperature rises are slightly higher than
those measured, perhaps indicating a slight difference in the actual heat rejected in the plant in
comparison with the values set by the boundary conditions in the model.
Analysis of theoretical condenser performance utilizing the methodology contained in the Heat
Exchanger Institute’s “Standards for Steam Surface Condensers” predicts summer/winter
pressures of 2.64/1.75 in Hg, suggesting that the condenser is performing satisfactorily and also
confirming the model predictions
90 10937790
4.3 Plant Variable Studies
Failed Scavenging System
In the runs performed to assess the failed scavenging system the outlet water boxes were
assumed to have a water level 18 inches below the top of the tubes. The flow rates to each
waterbox remained in the same proportion but were reduced in the affected tubes. Hence the
overall CW flow rate is reduced in these runs to 303,539 gpm, all other boundary conditions
were as for the corresponding base-case runs.
Effect of Air In-Leakage
The effect of air in-leakage was determined by increasing the prescribed value from 20 to 50
scfm. Hence, the air flow into the condenser was increased from 20 to 50 scfm by changing inlet
steam/air concentration to reflect this condition, and the air offtake pressure was increased by 0.2
in Hg from the corresponding base case runs to account for the effect of extra air loading on the
SJAE (Note: This result is crucially dependent on the response of the SJAE to increased air
loading.).
The results for these two studies are shown in Figure 9 and compared with the base-case results.
It can be seen that the magnitude of the effect of these latter influences is relatively small – the
air in-leakage having the most severe effect on pressure.
3.50
condenser pressure (inHg)
3.00
2.50 run 2-
run 3-
2.00 run 4-
1.50
1.00
30 40 50 60 70 80
CW temperature (F)
Figure 9: Graph of CW temperature versus Condenser Pressure for the base case (Run 2), failed
scavenging system (Run 3) and air in-leakage (Run 4)
Effect of Steam In-Leakage
10937790
91
Run 5 was carried out to study the effect of increased steam inflow into the West Shell. The runs
were carried out for inlet CW temperatures of 35, 65 and 75°F and for 1%, 2.5% and 5%
additional steam inflow. Figure 10 shows the effect on pressure for 35F and 65F CW water
temperatures.
1.58
1.56
1.54
pressure (inHg)
1.52 east
1.50 w est
1.48
1.46
1.44
1 2 3 4 5
steam in-leakage (%)
(a) CW Inlet Temperature=35F

2.70
2.68
2.66
pressure (inHg)
2.64 east
2.62 w est
2.60
2.58
2.56
1 2 3 4 5
steam in-leakage (%)
(b) CW Inlet Temperature=65F
Figure 10: Graph of Steam In-leakage versus condenser Pressure
The pressure in both shells rises, due to transfer of some of the steam to the East Shell, with the
West Shell pressure rising above that in the East Shell. The shell pressure differences rise to
about 0.05 in Hg at 5% steam in-leakage.
The CW temperature rises show slightly increased temperature in the West Shell, although the
differences are moderated by the cross flow of steam, which, as with the pressures in the shells,
tends to equalize the temperatures.
Plugged Tube Runs
These runs were performed to simulate the effect of tube plugging on condenser performance.
The East Shell tubes were plugged by 2%, 5%, 10%, 20% and 30% and runs carried out for inlet
water temperatures of 35, 65 and 75°F. Entergy has indicated that the East Shell tubesheet is
more likely to experience macro-fouling due to the layout of the CW piping.
92 10937790
The steam inflow and other boundary conditions remained the same as for the base case runs. As
tube plugging affects the overall cooling water flow, however, calculations were performed to
adjust the CW flows to each shell taking into account the flow curve and the frictional losses as
defined by the HEI Steam Surface Condenser Standards. The resulting CW flow boundary
conditions are given in Table 12 for the 35F inlet CW temperature case.
Tin plugged Qeast Qwest Qtotal

(F) tubes (gpm) (gpm) (gpm)
35 2 153767 156838 310605

35 5 151058 158955 310013
35 10 146385 162586 308971
35 20 136330 170360 306690
35 30 125237 178862 304099
Table 12: CW flow boundary conditions.
Figure 11 illustrates the effect of tube plugging on pressure. Values of 2% have little effect, but
as the percentage of plugged tubes increases the effect is more pronounced, with pressure
differences between shells rising to about 0.1 in Hg at 35F CW inlet temperature and to 0.16 in
Hg at 65F CW inlet temperature.
1.50
1.45
pressure (inHg)
east
1.40
w est
1.35
1.30
0 10 20 30
percent plugged
(a) CW Inlet Temperature=35F

2.65
2.60
pressure (inHg)
east
2.55
w est
2.50
2.45
0 10 20 30
percent plugged
Figure 11: Graph of percentage of East Shell Tube Plugging versus condenser Pressure
10937790
93
The temperature rises also follow a trend influenced by the CW flow rates. Temperatures in the
West Bundle fall slightly and in the East rise by several degrees. This behavior is perhaps closest
to that observed in the plant condition discussed earlier.
Figure 12 illustrates the effect of tube plugging on the CW temperature rises in each waterbox.
Tube plugging of 30% results in increases of about 3 to 6F in the temperature rises in the East
Shell and a 1 to 2F reduction in the West Shell. This is similar in magnitude to the changes in
CW temperature rises indicated by the Entergy plant conditions described in Section 1.1
36
34
temperature rise (F)
32 1--1
30
1--2
28
1--3
26
24 1--4
22
20
0 10 20 30
percent plugged
32.00
30.00
temperature rise (F)
1--1
28.00
1--2
26.00
1--3
24.00
1--4
22.00
20.00
0 10 20 30
percent plugged
Figure 12: Graph of percentage of East Shell Tube Plugging versus CW Temperature Rise
5 CONCLUDING REMARKS
This paper has reported the major highlights of a study carried out for Entergy’s Pilgrim Station. A three-
dimensional model of the Pilgrim condenser has been developed, utilizing CFD techniques, and applied
to simulate plant conditions. The model has been validated by comparing predictions for plant conditions
provided by Entergy for summer and winter conditions. Measured condenser pressure for summer
conditions was 2.6 in Hg (East Shell) compared with a prediction of 2.62 in Hg. The corresponding
winter conditions were 1.74 in Hg measured and 1.68 in Hg predicted.
94 10937790
Analysis of theoretical condenser performance utilizing the methodology contained in the Heat Exchanger
Institute’s “Standards for Steam Surface Condensers” predicts summer/winter pressures of 2.64 and 1.75
inches in Hg. This indicates that the condenser has a relatively good performance , a conclusion
confirmed by the modeling study. The similarity of measured, HEI and model predicted pressures indicate
successful validation of the model.
Following validation the model has been used to predict conditions for a range of inlet CW temperatures
and to study the effect on condenser pressure and outlet CW temperature rise of different possible plant
conditions. The main conclusions of these studies are as follows:
• The effect of the failed scavenging system is small, resulting in an increase of condenser pressure
of about 0.05 in Hg.
• Increasing the air in-leakage from 20 to 50 scfm results in a condenser pressure rise of about 0.1
in Hg.
• Increasing the steam in-leakage by 5% in the West Shell results in an increase of pressure of
about 0.05 in Hg in the East Shell and up to 0.1 in Hg in the West Shell
• The effect of 30% tube plugging in the East Shell, due to macro fouling, is to increase the East
Shell pressure 0.09 to 0.16 in Hg, the pressure increasing with increased inlet CW
temperature. The CW temperature rises are increased in the East Shell by 3 to 6F and
reduced by 1 to 2F in the West Shell.
These results provide a quantitative assessment of the effects of these conditions in isolation. It is
possible that the condenser is affected to an extent by all of the conditions studied. Further work
would be required to determine their additive effects on performance.
During the study assumptions have been made about the performance of the steam jet air ejectors
(SJAE’s). In particular, the effect of additional air in-leakage on the pressure set at the offtake by
the SJAE system. A better determination of the performance of the SJAE’s would be valuable.
As a first assessment, Entergy might wish to consider swapping SJAE’s, if possible, to assess
the effect upon condenser pressure.
10937790
95
10937790
EPRI Conference on Heat Rate and Cost Optimization
San Diego, Aug. 30-Sept. 1, 2005
The Cost of Dirty Condenser Tubes

A contribution to Condenser Plant Availability and Efficiency Improvements
By Remi P. Pullen, Manager USA, Technos, Alstom Group, remi.pullen@power.alstom.com

And Fred P. Lherminier, Engineering Consultant, Technos Consulting, TechnosCondenserWatch@terranov.com
Introduction
The deregulation of the US power generation market has pushed utilities and plants to improve
overall plant efficiency and performance. Projects that demonstrate short payback cycles and high
rates of return are singled out for capital expenditure.
However the gains from maintaining condenser cleanliness are often overlooked with little
correlation between cleanliness and output being made. A study by the authors using operating
data from many US plants indicates that the lost MWhs to cleanliness is phenomenal. Similar
studies performed on a number of sites are then compared.
With condenser cleanliness evolving permanently, continuous on-line tube cleaning and
monitoring is key to maintaining optimum performance. A discussion summarizes the
requirements for such a system.
Background
On-line Condenser Tube Cleaning Systems (CTCS) improve performance and protect the
condenser from fouling induced corrosion. They are not essential for the production of electricity
in the way that a boiler (steam generator), steam turbine and condenser are. Consequently at the
design phase where the cost of building the plant is the driving factor such systems are rarely
included.
Architect Engineers (AE’s) having won fixed price contracts are not responsible for performance
throughout the life of the plant and consequently will not specify non-essential equipment unless
required by the utility.
However we are seeing a change in existing plants where a number of plants are noting a
significant loss of performance and are starting to investigate causes and consider ways of
achieving or exceeding initial design performance.
It is interesting to note at this point that virtually all US plants have and are being designed with a
condenser cleanliness factor of 85%, while for a number of years all plants in Europe have been
10937790
97
designed with condensers to 95% clean. Unless a cleaning process maintains the cleanliness
factors, these remain design figures with no bearing on the reality. All power plant steam surface
condensers foul in operation. When first installed with new clean tubes they start their life with a
100% cleanliness factor (virtually never to be attained again) and rapidly start fouling with the
circulation of cooling water. The heating of the cooling water (CW) as it travels through the tubes
causes calcium carbonate and other deposits to settle out and form thick insulating scale on the
inside surface of the tubes. The condenser becomes a terrific feeding ground for microbiological
growth providing a warm environment with a continuous flow of nutrients. The end result is that
the condenser cleanliness factor can in reality drop fairly rapidly significantly below 85% and in
some cases in the region of 50% in particular with river or lake water CW.
Plants with condensers designed to 95% clean typically operate a reliable on-line CTCS which
maintains this cleanliness if operated correctly. Those designed to 85% clean typically rely on
outage off-line manual cleaning. Though if done properly the cleanliness reached after a good
manual scraping can be close to 100% it rapidly falls so that the cleanliness averaged over a year
is 70% or below. A number of studies on US plants using actual operating data collected over a
year confirm this scenario. The following is typical of our findings and is based on a real
900 MW unit of a US plant using actual site data.
Case Study
This 900 MW unit is coal fired and has a 3-shell triple pressure condenser. Using actual site data
logged over a full year CW inlet temperature versus condenser back pressure is plotted for each
shell and a trend line fitted to each plot. On the basis that the condenser was designed and built in
1974 the HEI of the day is used to plot a theoretical performance curve for each shell. By using a
cleanliness factor to trace the theoretical curve so that it matches the trend line the actual average
operating cleanliness factor can be concluded (See Figures 1, 2 &3 below). In the case the actual
operating condenser cleanliness factor is 63%. Also shown on these curves is the theoretical
performance at 95% clean, which would be obtained if an on-line CTCS were operated.
98 10937790
Figure 1
Figure 2
10937790
99
Figure 3
The heat load split across the LP, IP & HP shells is 38%, 29% & 33% instead of the design 35%,
30% & 35% given that the actual nominal CW flow is around 3.5% above the design value.
Using the guaranteed steam turbine backpressure correction curve supplied by the turbine
manufacturer and provided by the plant (See Figure 4 below) a correlation can be made between
MW electrical output and condenser performance. Inverting the axes of the condenser
performance curves and adding the turbine curve to the plot do this (See Figure 5 below).
Figure 4
100 10937790
Figure 5
Using average monthly inlet CW temperatures calculated from the data provided and capacity
and availability factors either provided or deduced, gains for each month can be ascertained. In
this study the following values were used: 4 cents per kWh, a capacity factor of 88% and an
availability factor of 95%. Table 1 below shows the gains on a monthly basis and the total gain
over the year, which equates to over 2 million dollars.
10937790
101
Return on Investment
Average CW Output Gain Operating Earnings
Month Output Gain Rate
temp * Period ** Gain
deg F kW h kWh $/kWh $
Jan 57 2 465 707 1 742 251 0.04 69 690
Feb 63 3 450 638 2 202 372 0.04 88 095
Mar 69 4 727 707 3 341 042 0.04 133 642
Apr 73 5 790 684 3 960 309 0.04 158 412
May 82 8 700 707 6 149 004 0.04 245 960
Jun 85 9 805 684 6 706 762 0.04 268 270
Jul 87 10 588 707 7 483 941 0.04 299 358
Aug 86 10 195 707 7 205 876 0.04 288 235
Sep 85 9 805 684 6 706 762 0.04 268 270
Oct 76 6 684 707 4 724 074 0.04 188 963
Nov 71 5 237 684 3 582 314 0.04 143 293
Dec 66 4 046 707 2 859 529 0.04 114 381
$ 2 266 569
* Incorporating Capacity factor of: 88%
** Incorporating Availability factor of: 95%
Table 1
The payback shown in Figure 6 below uses monthly gains from the table above and includes the
initial investment of $610 000 to purchase and install the required on-line CTCS.
Figure 6
102 10937790
Case Study Commentary
With this type of study the importance of monitoring condenser tube cleanliness throughout the
year becomes apparent. Once fouling sets in it is difficult to reverse without coming off-line.
Thus it is important to continuously monitor this calculated parameter and take corrective action
early on if it starts to fall. Plants that fail to do so may be able to live with the performance
shortfalls in the winter months but certainly feel the pinch in the warmer summer months often
having to de-rate in a period where cost of electricity soars.
The staggering gains demonstrated in the study are of course clearly in proportion to the MW
output rating of the plant. However based on similar studies carried out on a wide range of US
plants it is safe to say that virtually all power plants with an electrical output from steam of
250 MW and above would increase their earnings by at least several hundred thousand dollars a
year by retrofitting an on-line CTCS and that the payback for such a system would be in most
cases well under a year.
Multiple site analysis

Table 2 below summarizes the design and operational characteristics of 4 different condensing
units, and table 3 shows results of the analyses performed with the method described above.
Site # Nominal Heat Nominal Tube Gauge No of No of tubes Tube Base electric
load CW Flow diameter (BWG) pass per pass length load (MWe)
(MBtu/hr) (gpm) (Ft)
1 3 430 000 227 900 0.875 18 2 22 040 41 640
2 1 507 870 292 683 1 18 1 17 726 28 900
3 3 561 658 360 900 1 18 1 24 152 50 550
4 2 699 500 330 000 1.25 22 2 11 272 40 650
Table 2
Site Calculated Target Averaged Peak Monthly Average Payback 24

# cleanliness cleanliness power power output gain net period months
factor factor with gain gain (MWh, 12 monthly (months) revenue
CTCS (MWe) (MWe) month cash flow increase
average)
1 76% 95% 11.6 23.9 1287 $ 51 500 9 k$ 740
2 63% 95% 6.8 10.5 4722 $ 188 900 5 k$ 3 920
3 76% 95% 1.5 3.8 1075 $ 64 500 6 k$ 1 350
4 70% 95% 1.3 3.3 931 $ 37 200 19 k$ 220
Table 3
The payback period (which accounts for CTCS investment) is less than one year, except for site 4
where cooling water conditions are less favorable. In all cases a significant positive cash flow is
generated.
10937790
103
On-line Monitoring, Modeling, Diagnosis and Prevention
In today’s deregulated market where staffing levels at plants have been reduced just retrofitting a
CTCS is not sufficient if its operation is not monitored.
Considering the economical benefits of improving and maintaining high condenser performance
under varying operational conditions (such as cooling water characteristics), the need for a real
time performance monitoring, analysis and decision tool is clear. Such a tool would also be
extremely beneficial to another key requirement of power plants: availability. Finally, many
countries now impose constraints on CO2 releases as well as CW outlet temperature limits. These
can prove to become bottlenecks to power production (by limiting allowed primary source
consumption, or by restraining the availability of the cooling source).
The design was therefore oriented toward these two goals: performance and availability, through
monitoring, feedback, and prevention.
Below is a summarized functional description of the implemented solution along with a synopsis
of embedded modeling, features and operational benefits to the plant owner.
Overview of requirements
The main purpose of the developed tool was to provide the plant staff with day-in day-out
intuitive decision helping tools aimed at optimal condenser operation and easy preventative
maintenance scheduling. Hence the tool was to be installed on site, with real-time measurements
and state-of-the-art networking and communication tools. This resulted in an extremely versatile
data acquisition and automated report dispatching platform.
The second objective was to maximize such benefits without implying any kind of complex and
expensive instrumentation scheme. Therefore the tool had to be capable of integrating real time
data from a variety of existing sources, and avoiding additional instrumentation where possible.
Namely, we restricted the necessary instrumentation to the following sensors:
• Cooling Water Inlet and Outlet Temperatures,
• Cooling Water Flow (obtained from Condenser Pressure Differential)
• Condenser Vacuum Pressure.
Additionally, we required the tool to offer compatibility with DCS communication protocols in
order to easily accommodate reuse of existing data commonly available (e.g. instantaneous
electric power, condensate temperature, and more). Where necessary (e.g. multi body
condensers), the instrumentation scheme may be duplicated. The server is sized to accommodate
up to 16 condensing units.
The next requirement was to be able to embed a community approved real time modeling tool, as
well as custom developed decision logic. The HEI 9th edition was selected as being the most
recognized condenser performance analysis standard. The model was also matched with in-house
design tools so that initial design and commissioning conditions can be input as model reference,
including design factor.
In order to allow for preventative maintenance schemes, additional criteria were derived,
including water chemistry and statistical signal analyses (trends). Coupled with the embedded
104 10937790
logic engine, this set of helping hand tools makes the most of the available data. They dictate
actions to be taken in order to maintain the highest levels of performance while minimizing
aging.
Figure 7 - Condenser Watch Bloc Diagram

Design principles
In a fully featured version, the Condenser Watch incorporates:

• Real time condenser performance assessment and reporting
• Economical impact analysis
• Online analysis of cooling water chemical properties (pH, alkalinity)
• Online diagnosis of condenser and auxiliary devices combined operation
• Dynamic scheduling of preventative maintenance operations
• Alarms and events logging and management
• Data logging with user friendly data viewing tools
According to the specifics of each site, from condenser design constraints to available
measurements and data, the Condenser Watch features are activated or inhibited to best suit
operational reality. Actual condenser design parameters (layout, flow patterns, dimensions, and
materials) are entered into the embedded model. Adjustable alarm thresholds and custom
automated reporting allow the operators to manage the level of feedback the system provides
them. Alarm notifications, system status reports and historical data logs may be transmitted by e-
mail with both event trigger and regular interval scheduling.
10937790
105
Figure 8 - Condenser Watch Supervisory Interface sample
Figure 9 - Condenser Watch Alarms, events and diagnosis tools sample
Conclusion
A number of detailed analyses have been performed based on actual condenser data in various
configurations, establishing in each case the huge achievable gains and savings offered by steady
on-line CTCS condenser cleaning.
The benefits of such readily available condenser cleaning technology are all the more apparent
when coupled to the Condenser Watch, an intuitive and purpose oriented monitoring, on-line
analysis and condenser diagnosis tool.
Further documentation is available online at http://www.technos.fr

(Product brochure may be downloaded at http://www.technos.fr/2004/PDF/Condwatch.pdf)
106 10937790
3
SESSION 3: MAINTENANCE TECHNOLOGY I
10937790
107
10937790
Eddy Current Application for
Balance-of-Plant Heat
Exchangers
Nathan Muthu
Project Manager
EPRI NDE Center –BOP Division
Charlotte, NC 28262
Tel: + 704.547.6046
Email: nmuthu@epri.com
10937790
109
Common Types of Tubing Material Used
in Heat Exchangers
• Nonferromagnetic
• 300 Series Stainless Steel (304 and 316)
• Higher Grade of Stainless Steel (AL6XN, AL6X)
• Yellow Metals (Admiralty Brass, 90/10 CuNi,
Aluminum Brass, Copper – Prime Surface, and
Finned)
• Titanium
• Ferromagnetic
• Monel 400 series
• 439 Stainless Steel
• SeaCure
• Carbon Steel
© 2005 Electric Power Research Institute, Inc. All rights reserved. 2
110 10937790
Tubing Configuration
• Prime Surface Tubing

– Smooth OD and ID Surfaces
• Integral Finned Tubing
– Finned OD with Smooth ID Surface
• Wrapped Finned Tubing
– Finns Wrapped Around OD of Prime Surface Tube
• Solder Finned Tubing
– Finns Soldered to OD of Prime Surface Tube
10937790
111
Typical Types of Damage Mechanism
Seen in Heat Exchanger Tubular Products
• ID Pitting (Corrosion and Microbiological Influenced

Corrosion – MIC)
• ID Wall Thinning caused by erosion
• OD Circumferential Cracking
• OD Axial Cracking
• OD Steam Erosion
• Tube-to-Tube Support Plate Wear
• Tube-to-Tube Wear
• Loose Part Damage
112 10937790
Typical Types of Damage Mechanism Seen in
Heat Exchanger Tubular Products …cont’d
OD Circumferential Cracking
OD Axial Cracking
OD Steam Erosion & Hole
ID Erosion
10937790
113
Tube- to-Tube Support Plate Wear and

Tube- to-Tube Wear
Through Wall Hole
Condensate Groove under TSP Location Steam Erosion in 70-30 CuNi
114 10937790
Pits in Tube ID – Finned External

Local Pits in OD Freespan Section
Large Pits from Chemical Attack

Ammonia Groove in Brass Tubing
10937790
115
OD Axial Cracking
Underneath TSP in 304
SS Condenser Tube
1.00” (25.4mm) x 0.028”
(0.711mm)
Same axial crack seen in

the dissected Tube – ID
Note the transverse
branching of the crack
116 10937790
OD Axial Crack & Branching of OD Circumferential Cracking

on 304 SS, 1.00” OD x 0.028” (25.4mm x 0.711mm) found in
main steam condenser tube – under tube support structure
10937790
117
Microbiological
Influenced
Corrosion
118 10937790
Destructive Sectioning of Chloride Pitting in 304 SS Tube
Step 1 Step 2
Step 3 Step 4
10937790
119
Tube Support Plate Damage
Complete Loss of Tube Support Plate Member in

Low Pressure Feedwater Heater
Tubes: 304 SS , 0.750” OD x 0.049” (19.05mm x 1.24mm)

TSP: Carbon Steel, 0.625” (15.875mm) nominal thickness
120 10937790
Complete Loss of Tube Support Plate Member in

Low Pressure Feedwater Heater
10937790
121
Complete Loss of Tube Support Plate Member and Structural

Damage in Low Pressure Feedwater Heater
122 10937790
Eddy Current Examination……
A very good “Preventative Maintenance” Tool

Provides Repeatable & Reliable Results
Affords the utility to take timely measures to plan for repair
and replacement (R&R) activities
Prevent Emergency Plant Shutdowns Caused by Tube Leaks
Allows the operator to modify water chemistry, operation,
and design parameters based on damage mechanism found
Allows Data Trending to Monitor Damage Growth Rates
Offers a Safe, Reliable and Extended Operation of Heat
Exchangers
$$$ Savings – R&R Activities/Emergency Plant Shutdowns
10937790
123
Keys to a Successful Eddy Current
Examination
• Inference based - Use appropriate calibration
standards
• Use the correct probes
• Selection and application of appropriate
frequencies and channels
• Use multi-frequency eddy current equipment
• Prior history of damage mechanism, if available
• Pulled tube destructive results, if available
• Previous eddy current test results, if available
• Qualified Data Analyst ***
124 10937790
Components of a Successful Eddy
Current Examination
ET Error
Test Frequencies
10%
15%
History / Verification
10%
TSP Mixes
10%
Qualified Analyst
25% Calibration
Standards
15%
Analysis Techniques
15%
10937790
125
Eddy Current Multi-Frequency Equipment
Zetec Miz-27/SI CoreStar Omni 100
R/D Tech TC-5700
126 10937790
Eddy Current Probes
10937790
127
Eddy Current Calibration Standards
128 10937790
Advantages and Limitations
• Advantages
– Fast Production Speed >> 12”/Second
– No Physical Couplant Required
– Provide Flaw Location, Origin, and
Percent Wall Loss
– Reliable & Repeatable Data
– Comparable Data for Trending &
Growth Rate
• Limitations
– Affected by Both Relevant & Non-
Relevant Tubing Conditions
– Experienced Analysts Required
10937790
129
Eddy Current Analysis Guidelines
• General Guidelines for Identifying Flaw Signals

– All Flaw Signals without Interfering Non-relevant Signals
Occupy the Flaw Planes at All Operating Frequencies,
i.e., High, Intermediate and Low
– For a Given OD Flaw, Measured Phase Angles Decrease
from High to Low Operating Frequencies
– For a Given OD Flaw, Signal Amplitudes Generally
Increase with Decreasing Operating Frequencies
– For an ID Flaw, Measured Phase Angles Remain within
an ID Phase Plane but Increase from High to Low
Frequencies
– For an ID Flaw, Signal Amplitudes Generally Decrease
with Decreasing Operating Frequencies
130 10937790
EPRI-Recommended Calibration Set-ups

•TWH from ASME Calibration Standard set
Individually to 40 degrees at 5.00 volts at all
differential channels
5.00 volts 5.00 volts 5.00 volts 5.00 volts

p-p p-p p-p p-p
40-deg 40-deg 40-deg 40-deg
Ch.1 Differential Ch.3 Differential Ch.5 Differential Ch.7 Differential

High Frequency High intermediate Intermediate Low Frequency
Frequency Frequency
10937790
131
20%
40%
60%
80%
Direction of Probe Pull 100%
TSP
ASME Type Calibration Standard Ring
Eddy Current Bobbin Probe
132 10937790
Lissajous Signal Amplitude Curve
100
ID Flaw OD Flaw Plane
90
Plane 1/16” Dia. Pits
80 3/32” Dia. Pits
70
% Wall Loss
60
50
40
30
1/8” Dia. Pits
20
10
20 40 60 80 100 120 140 160

Phase Angle
Round Bottom ID Pits Elongated ID Pits
Varying Diameter and Types of Pits EDM on Tube ID

10937790
133
Microbiologically Influenced
Corrosion (MIC)
• Primarily affects Carbon Steel, Copper Alloys and
Stainless Steels (300 and 400 series)
• Material factors which contributes to MIC are:
- Carbon Steel – Gouges, dings, inclusions, defective
millscale surfaces, and welds HAZ’s of millscaled pipe
- Stainless Steel - Gouges, dings, inclusions, heat tint,
defective welds, and sensitization
• Silt and other suspended solids are primary carriers of
bacteria into the system. Consequently, low/static flow
rate is a primary contributor to MIC since silt and other
suspended solids can settle on component surface and
promote MIC initiation
• Other low/static flow rate influences are lack of adequate
biocidal treatment and dissolved oxygen supply
134 10937790
MIC Flaw Signatures
• MIC typically originates on the tube ID side, caverns

through the tube wall thickness and may exit on the
tube OD side
• As a result of the small ID openings coupled with
cavernous mid-wall loss extending to the tube OD, the
eddy current flaw signals form more on the OD plane. Tube OD
Cavern Effects
10937790
135
MIC Flaw Signatures
• By using Multi-Frequency, Multi Channel Inspection

Techniques with the EPRI-recommended calibration setups,
the lowest frequency channel (Channel 7) will cause the ID
originating MIC signal to fall into the ID plane. This is one
way to confirm the presence of MIC damage form.
OD Plane OD Plane ID/OD Plane ID Plane
52-deg 48-deg 40-deg 32-deg
Ch.1 Differential Ch.3 Differential Ch.5 Differential Ch.7 Differential

600 kHz 300 kHz 150 kHz 75 kHz
136 10937790
MIC Flaw Sizing
• When dealing with MIC damage, sizing flaws using phase

analysis alone may be extremely conservative – end up
plugging more tubes.
• Suggested method to evaluate MIC:
- retrieve a tube from service
- identify the flaw morphology
- obtain ground truth results
- make appropriate calibration standards
- use both Phase and Amplitude information to determine
tube plugging criteria.
10937790
137
Controls for MIC
• Clean and maintain dry component

surfaces
• Flush and place the tube waterside into
forced air dry lay-up
• Use of biocides, e.g., chlorine or bromine
for treatment of the circulating water or
other raw water cooled systems
138 10937790
Eddy Current Performance
Demonstration Tests
• Use Qualified Techniques & Analysts

– Reliance on Realistic Mockups
– Reliance on Damage-Specific Cal Standards
– Continuous Training
– Experience
10937790
139
Qualified Analyst- Key to Successful
Eddy Current Program
• Ability to Discriminate Relevant Flaws from

Non-Relevant Indications
• Ability to Depth Size Flaw Indications using
Appropriate Calibration Standards and Calibration
Curves
• Desirable to have Analysts Successfully Complete
Site-Specific Performance Demonstration Tests
– Relies on Realistic Test Samples
– Contains Flaw Types Expected in the Field
– Passes both Flaw Detection & Flaw Sizing
140 10937790
BOP Heat Exchanger Performance
Demonstration Test
• Rely on Realistic, Well-Characterized Air

Transportable Mockups
• Evaluate Proficiency of Data Analysts to Detect &
Size Flaws
– 80% Passing Score Needed for Detection & Sizing
• Conducted Either at the Vendor Site or at the EPRI
NDE Center
• On-going Performance Demonstration for
– Power Industry
– Chemical & Processing Industry
10937790
141
EPRI Air-Transportable Mockup Bundles
• Extensive Collection of Realistic Mockups

– Nonmagnetic Tubing - Admiralty Brass,
Aluminum Brass, 90-10 Copper Nickel, Type 304
Stainless Steel, Titanium
– Magnetic Tubing - SA210 Carbon Steel, SA179
Carbon Steel, SA556 Carbon Steel, 439 Ferritic
Steel, Monel 400
Currently 15 Full-Size &
17 Air-Transportable
Mockups Available at the
EPRI NDE Center
142 10937790
Ferromagnetic Tubing Inspection
• Carbon Steel
• Monel 400
• Thin Wall High Performance SS such as SeaCure®
and 439 SS
Inspection Methodologies
• Remote Field
• Magnetic Saturation
• Partial Saturation
• Full Saturation
• Flux Leakage
10937790
143
Ferromagnetic Tubing Inspection
• Generally single frequency is used for ferromagnetic

tubing inspection (Carbon Steel,SeaCure®, and 439
SS)
• Monel 400 – Strong rare earth magnets allow Monel to
behave as non-magnetic tubing:
– Good phase angle separation between 100% and
20% through-wall holes
– Magnetic-biased multi-frequency ET testing
• If Monel tubing is not magnetically saturated (no
phase angle separation), apply the following:
– Partial-saturation ET technique
– Remote-field ET
144 10937790
Trending / Forecasting
• Objectives
– Estimate Damage Growth Rate
– Predict Remaining HX Life
• Methodology
– Review & Evaluate Successive Eddy Current
Examination Results
– Examination Results Must be Comparable
• Essential Examination Variables Must be
Maintained from One Outage to Next
– Calculate Remaining HX Life
10937790
145
Remaining Life Calculation
• Rely on Two Comparable Data Sets – Same Essential

Variables Used to Acquire ET Data Sets
• Calculate Growth Rate - Perform Regression Analysis
of Two Comparable Data Sets for Flaw Growth Model
– Exported to Microsoft Excel for Trending and
Estimation of Flaw Growth Rate
– Assumptions Needed for Remaining Life Estimate
• Known Plugging Criterion, I.e., 40%, 60%, 70%,
etc
• No. of Plugged Tubes to Date
• No. of Allowable Tubes to be Plugged
146 10937790
Effective Heat Exchanger Inspection
Program Can:
• Reduce Maintenance Cost

– Minimize Tube Leaks - Effective Inspection Plan
– Establish Realistic Plugging Criteria
• Increase Availability of Heat Exchangers
– Extend Examination Interval
– Reduce Insurance Plugs
• Provide Realistic HX Assessment
– Provide No. of Plugged/Degraded Tubes
– Estimate Growth Rate of Observed Damage Form
– Predict Remaining Operating Life
10937790
147
Summary
• Reliable Eddy Current Test Results Can Help Improve

the Effectiveness of Heat Exchanger Inspection and
Maintenance Practices
• Qualified Analysts Integral Part of Optimized
Inspection Process
• Consistent and Comparable Eddy Current Data
Required to Estimate Current HX Conditions and
Remaining HX Life
148 10937790
EPRI Utility Assistance Related to BOP
Heat Exchanger Inspection Activities
• Eddy Current Theory and Application Training Pertaining to Balance

of Plant Heat Exchanger Tubular Inspection
• Specific Data Analysis Training
• Technical Review of Field Data
• Procedure Preparation and Qualification
• Surface Eddy Current Application
• Remote Field Applications for Ferromagnetic Tubing
• Heat Exchanger Remaining Life Calculation
• Site Specific Performance Demonstration Development Using Utility
Component Specific Data
• Probe Performance Evaluation
• BOP Heat Exchanger Performance Demonstration Testing
• Fabrication of Calibration Standards, Mock-Up Test Bundles
• Biennial BOP Heat Exchanger Symposium – Big Sky, MT, June 2006
• Other Utility-Requested Assistance
10937790
149
10937790
RFT Inspection Techniques to
Characterize and Size OD Cracks
at TSP Location for Thin Walled
High Performance Ferritic
Stainless Steel Tubing
Nathan Muthu
Project Manager
EPRI NDE Center –BOP Division
Charlotte, NC 28262
Tel: + 704.547.6046
Email: nmuthu@epri.com
10937790
151
Overview
• Crack detection directly underneath external structure such as

TSP proves to be a formidable challenge for RFT Inspection or
Magnetic Saturation techniques.
• Progress Energy experienced axial cracking in SeaCure®
tubing material in the main condenser.
• Crack was noticed only because it extended beyond the TSP
width on the tube using a magnetic saturation probe and
examination technique.
• Tube Dimension: 1.00”OD x 0.028” (25.4mm x 0.711mm)
• TSP Width: 0.750” (19.05mm)
• Flaw Length: ~ 1.125” (28.6mm)
• Tube in-service for approximately 18 months
• Funding obtained from four NDE Center Utility Members
152 10937790
Flaw and Signal Representation
Carbon Steel TSP Region TSP Signal
Tubing Flaw
Signal
Axial Crack and Other Small Volume Flaws Under

Carbon Steel Support Plates Offers a Formidable
TSP Signals are large due to the
Challenge for Detection and Characterization magnetic field absorption,
therefore masking tubing
flaws
10937790
153
Collaborative Partners and Vendor
• Initial Collaborative Partners in 2003 & 2004:

– Progress Energy
– Exelon Nuclear
– Tennessee Valley Authority (TVA)
– Electricite de France (EDF)
• 2005 Supplemental Funded Project by NDE Center Members
• Sensor Manufacturer’s
– Carbon Steel Inspection, Pittsburgh, PA - completed
– Russell NDE Systems, Edmonton, Canada- completed
– Zimmerman Technical Services, Seattle, WA - completed
– Eddy Current Technologies, Virginia Beach, VA- completed
– CoreStar International, Pittsburgh, PA – in progress
154 10937790
Air-Transportable Mock-Up Bundle
• EPRI Fabricated Air-Transportable Mock-Up Bundle

Consisting of SeaCure® Tubing, 1.00” OD x 0.028”
and 0.035” wall (25.4mm x 0.711mm and 0.889mm) to
assist in probe development and R&D.
Fabricated Flaws Included:
• OD / ID Pits
• OD Axial and OD Circumferential Notches (Free span
and TSP location)
• OD Wall Thinning and Erosion
• Wear Scars Underneath Tube Support Plates
10937790
155
Carbon Steel Inspection (CSI), Pittsburgh, PA
6-Channel Switchable Array Probe
Carbon Steel Inspection (CSI) manufactured one

0.870-inch (22.1mm) diameter 6 channel switchable
RFT array probe (85% fill-factor).
4
Front 1 Rear
Exciter 5 Exciter
2
6
Probe Can be Used in Three Different Modes:

•Front Single Exciter
•Rear Single Exciter
•Dual Exciter
156 10937790
Carbon Steel Inspection (CSI), Pittsburgh, PA
External Reference Remote Field Probe
Test Probe
Exciter Exciter
Tube
Test
Equipment
XRFT Reference Probe

Inside In-Situ Tube Bundle
Sensor A Sensor B
10937790
157
Benefits
• Overcomes the depth sizing limitation of the general
purpose RFT bobbin probe and can be used in three
different modes; Front, Rear, and Dual Exciter modes
• Radial offset of the three coils 60º from the other coils
allows the indication or flaw geometry to be established
(flaw profilometry)
• Provides less volumetric dependency because the
coils do not average over the entire circumference like
the bobbin coil
158 10937790
Disadvantages when Operated in Dual Exciter Mode

•Shallow small volume flaws (up to 60%) in tube mid-span
and all small volume flaws regardless of depth are not
detectable directly underneath or adjacent to support
structures when operated in dual exciter mode.
When Operated in the Single-Exciter Mode:
• Helps to increase the sensitivity to small volume flaws
located adjacent to an external structure but not directly
underneath.
Equipment Requirement:
•Need two RFT Modules when using Zetec’s Miz-27/SI
test equipment to operate six channels.
10937790
159
6-Channel Switchable Array Probe – Dual
Exciter Nominal TSP Signal
Coil Coil
1 2
Coil Coil
3 4
Coil Coil
5 6
160 10937790
Exciter
75% 180º OD Wear Scar
Coil 1 Coil 2
Coil 3 Coil 4
Coil 5 Coil 6
10937790
161
Exciter 50% 180º OD Wear Scar & 50% Axial Notch under
TSP – only wear scar can be identified and sized
Coil 1 Coil 2
Coil 3 Coil 4
Coil 5 Coil 6
162 10937790
Exciter 100%, Axial Notch- 1.125” (28.75mm) long extends
beyond TSP width into Free Span region
Coil 1 Coil 2
Coil 3 Coil 4
Coil 5 Coil 6
Distorted Signatures
10937790
163
Exciter 75%, Axial Notch- 1.125” (28.75mm) long extends
Coil 1 Coil 2
Coil 3 Coil 4
Coil 5 Coil 6
Axial Notch Cannot Be Discerned
164 10937790
6-Channel Switchable Array Probe –
Front Single Exciter
Nominal TSP Signal
Coil 1 Coil 2
Coil 3 Coil 4
Coil 5 Coil 6
10937790
165
6-Channel Switchable Array Probe – Front
Single Exciter 100%, Axial Notch- 1.125” (28.75mm) long extends
Coil 1 Coil 2
Coil 3 Coil 4
Coil 5 Coil 6
Distorted Signals In All Channels
166 10937790
6-Channel Switchable Array Probe –
Front Single Exciter 75%, Axial Notch- 1.125” (28.75mm) long
extends beyond TSP width into Free
Span Region
Coil 1 Coil 2
Coil 3 Coil 4
Coil 5 Coil 6
Cannot be Discerned
10937790
167
Exciter (Field Trial)
Coil 1 Coil 2
Coil 3 Coil 4
Coil 5 Coil 6
168 10937790
Exciter Mode – Flaw Voltage Response
(Field Trial)
Voltage Response Per Array Coil At Defect
0.01 Volts 0.06 Volts
0.14 Volts
0.14 Volts
0.64 Volts
Coils
1 2
3 4
0.06 Volts
5 6
10937790
169
Flaw Profilometry Offered by Array Coil Design
(Field Trial)
Flaw Profilometry As Seen By Each Array Coil
Coil 1
Coil 6 Coil 2
Coil 5 Coil 3
Coil 4
170 10937790
RESULTS – Flaw Type Detection Dispersion
Carbon Steel Inspection
110
100
OD and ID Pits
90
80 180 and 360 W /Scar
Percentage Thru Wall
70
Axial Notch in F/Span
60
50 90 and 180 Circ Notch in
F/Span
40
Steam Erosion
30
20 OD Thinning
10
0
1 2 3 4 5
Number of Flaw s
10937790
171
Russell NDE Systems –Edmonton, Canada
Saturn® Probe
• Saturn® sensor design is operated in the “Near-Field Eddy Current”

low-frequency spectrum of 20kHz to 30 kHz.
• Standard Saturn® sensor is designed with one exciter and two
detector coils arranged in close proximity between one another
offering the near-field eddy current mode.
• Pancake coil design consists of 6 coils arranged in one bracelet.
• Both coil designs were tested with and without magnets.
• Magnets were used to partially saturate the tube wall (align the
magnetic domains) so that free-span tubing flaws could be identified
and sized accordingly.
• Flaws underneath tube support plates were investigated in the near-
field eddy current region using low frequency without magnets
• Pancake coils were used to enhance detection of axial (longitudinal)
flaws. Also used in providing flaw profilometry and location.
172 10937790
Russell NDE Systems –Edmonton, Canada
Finite Element Simulation (FES)
• Used in conjunction with the near-field eddy current

techniques to simulate anomalous source signals.
• Anomalous source algorithm is required to extract defect
signal by eliminating any other background signal such as
the tube support plate (TSP) signal.
• Once the TSP signal is eliminated, only the tubing flaw
signal is present for evaluation.
• In principle, the anomalous source algorithm is useful for
identifying defects under TSP.
• Big achievement and success today. Typically,

tubing flaws under TSP cannot be characterized. This has
been an accepted by the industry as a test limitation up
until today. The hurdle has been crossed successfully.
10937790
173
Simulated Anomalous Groove Signal
X-Y Plot
Strip Chart Plot

Shown here is the end result after subtracting the simulated TSP signal
from the combined measured TSP signal and simulated groove signal.
174 10937790
Defects in Free Span Tubing Area Detected with
Standard Saturn® Probe
10937790
175
Defects in Free Span Tubing Area Detected with
Multi-Channel Pancake Saturn® Probe
176 10937790
Anomalous Source Signal for Defects Underneath
Tube Support Plates
10937790
177
RESULTS – Flaw Type Detection Dispersion
Russell NDE Systems
100
90 OD and ID Pits
80
180 and 360 W /Scar
70
Percentage Thru Wall
60 Axial Notch in F/Span and

under TSP
50
90 and 180 Circ Notch in
40 F/Span and under TSP
30 Steam Erosion
20 OD Thinning
10
0
1 2 3 4 5
Number of Flaw s
178 10937790
Conclusions
• Saturn® probes are highly sensitive to all types of

defects machined in SeaCure® tubes.
• Major defects underneath Tube Support Plates can be
identified and detected.
• Several sensor techniques and data interpretation
methodologies can be combined in Saturn probes.
• Major Breakthrough in the Nondestructive Testing
Arena for Ferromagnetic Tubular Product Examination
10937790
179
10937790
RETUBING
LARGE STEAM SURFACE
CONDENSERS
ACTIVITIES, SPECIALIZED TOOLS,
and PROCEDURES
Ed Overmyer
Mark Tagliamonte
RETUBECO, INC.
www.retubeco.com
10937790
181
ABSTRACT
A typical steam surface condenser retubing
project involves removing and installing
hundreds of tons of material (the tubes) a few
pounds at a time.
This presentation is an overview of activities,
specialized tools, techniques, and general
procedures necessary to efficiently and
properly retube a condenser
The operation of specialized tools for
internally cutting tubes, freeing the tube to
tube sheet joints, flattening and removing old
tubes, and chopping tubes into short lengths
for disposal are illustrated.
Tube sheet and support plate hole
preparation are shown, as are tube insertion,
expanding, trimming, and flaring.
182 10937790
TYPICAL TASKS and SCHEDULE
10937790
183
TYPICAL STAFFING &
EQUIPMENT FOR 24,000
TUBE PROJECT
Schedule 2 shifts/day x 10 hrs/shift x 6 days/wk
Duration 24 work days ± 24-30 calendar days
1 Project Manager + 1 Second Shift lead
QC coverage
2 Technicians per shift
20-25 Craft per shift
4 Complete sets of tube removal equipment
8-12 Sets of support plate cleaning equipment
8 Sets of tube sheet cleaning & rolling
equipment
184 10937790
CRITICAL PATH
ACTIVITIES TO MONITOR
% Man/hrs % Shifts
Tube removal 15-20 17-19
Support plate cleaning 5 -10 6±8
Tube installation 40-45 19-22
Total % of project 60-75 42-49
10937790
185
EQUIPMENT STAGING
186 10937790
TUBE REMOVAL
Gain water box & steam side access
Tube plug removal, both ends
Internal tube cutting, back tube sheets
Tube stub removal, back tube sheets
Water box removal, front tube sheets
Erect scaffolding, front tube sheets, 6-8 ft.
work platform width
Install rigging for tube removal
Pull, travel, and chop tubes, front tube sheets
10937790
187
CUT TUBES, BACK SIDE
188 10937790
PULL STUBS, BACK SIDE
10937790
189
WATERBOX REMOVAL,
FRONT
190 10937790
PULLING TUBES
10937790
191
Tube Traveler Operation
4 Travelers ² 2 Tubesheets
192 10937790
TRAVEL TUBES
10937790
193
TRAVEL & CHOP TUBES AT
THE TUBE SHEET
194 10937790
CHOPPING TUBES
10937790
195
SCRAP TUBE
Typical Removal Rates 2-3 Tons/hr.
196 10937790
TUBE INSTALLATION
Ball burr support plate holes
Check support plate alignment
Clean condenser steam space
Brush clean tube sheet holes
Inspect & repair tubesheet holes
Install new tubes
Roller expand tubes
Trim and bell tube ends if required
10937790
197
SUPPORT PLATE HOLE
CLEANING
Critical activity to aid tube sticking
A large condenser will have several
hundred thousand holes to clean
Most applications use carbide ball burrs
0.015´ +0´/-0.002´ larger than tube O.D.
Burrs are driven by 1 hp, 400 rpm drill
motors
Production rate is several seconds per hole
Check holes with Go/No Go gauges
Ball burring does not remove metal from
holes
198 10937790
TOOLS TO CLEAN
SUPPORT PLATE HOLES
10937790
199
TUBE SHEET HOLE
CLEANING
Depending on tube sheet material
brushing or abrasive flapper
wheels may be used to clean tube
sheet holes
After cleaning tube sheet holes
are inspected for defects
Damaged holes are marked for
repairs
200 10937790
TOOLS TO CLEAN
TUBE SHEET HOLES
10937790
201
SERRATE TUBE SHEET HOLES
Tube sheet holes can be serrated with 8 point bits for
additional joint strength.
202 10937790
TUBE SHEET HOLE REPAIR
Flapper wheel to remove light
longitudinal indications
Roller burnishing tools to remove
deeper longitudinal indications,
ovality, hole taper, and burrs
Two part epoxies for filling gouges
Reaming and bushing installation
for major defects
10937790
203
NEW TUBE
INSTALLATION
Inspection hold point after all cleaning
and repairs to tube sheet & support
plate holes
Tubes are staged to the sticking deck
and tube guides installed
Tubes are installed top down (typical)
If tube stakes are required they may be
installed in parallel with tube installation
depending on configuration
204 10937790
ANTI-VIBRATION STAKING
Thinner wall tube may require vibration analysis
and mid span staking.
10937790
205
PUSH OUT TEST
Testing can be performed to determine the optimal tube
to tube sheet joint configuration given the tube and
tube sheet thicknesses and materials.
206 10937790
TUBE STICKING
Tube sticking is much more productive when support plate
alignment has been checked and all holes cleaned with
specially sized carbide ball burrs. Tube sheet holes are
brushed clean with double twist stainless steel hole
brushes, 1/16´ larger diameter than the tube hole.
10937790
207
TUBE ROLLING
Procedure, craft training, and monitoring
are critical
Usually inlet tube sheets rolled in first
Then a tacking pattern is rolled on outlet
tube sheets prior to production rolling
Tubes are set and roller expanded using
the known hole method
Typically every 100 hole is used as a
benchmark
Torque controlled rolling motors are
checked for consistency with Cleco P-15
torque analyzers
Rolling motor air supply is also monitored
during rolling operations
208 10937790
TUBE ROLLING TOOLS
Airetool 850-600 Torque Controlled
Rolling Motor, 5 Roll Expanders
TUBE ROLLING
Airetool 850-600 Torque Controlled
Rolling Motor
3 Roll and 5 Roll Expanders
10937790
209
TUBE ROLLING
210 10937790
TRIMMING & FLARING
If required trimming and flaring
can be performed concurrently
These activities are performed
after both ends of the tube have
been roller expanded into the tube
sheets
10937790
211
TUBE TRIMMING
212 10937790
TUBE FLARING
10937790
213
TYPICAL TUBE INSTALLATION
214 10937790
FINAL TESTING STEAM SIDE
Install water boxes prior to test (design)
Fill water should be above 65°F
Fill rate is typically 1/2´/minute after
water reaches tube bundle
Fill to expansion joint or 3 feet above
bundle (dye added after fill complete)
During fill inspect tube to tube sheet
joints for leaks & re-roll as required ±
stop fill if necessary to support re-roll
Typical hold period 12 hrs after all leaks
stopped
10937790
215
FINAL TESTING CW SIDE
Follow normal plant operating
procedures
Fill water boxes and perform in-
service test
Inspect water box gaskets for
leaks, retorque bolts as necessary
Hold time is plant procedure
specific
216 10937790
SEPTEMBER 2001
10937790
217
10937790
IMPROVING CONDENSER EFFICIENCY WITH
INNOVATIVE SCALE REMOVAL SYSTEM TECHNOLOGY
Jon T. Hansen George Saxon, Jr.

Omaha Public Power District Conco Systems, Inc.
North Omaha Station 530 Jones Street
7475 Pershing Drive Verona, PA 15147 USA
Omaha, NE 68112 USA gsaxonjr@concosystems.net
jhansen@oppd.com
Abstract
Identified as one of the single largest energy loss factors in power generation stations,
condensers are a potential source of either savings or increased expenditure, especially as fuel
prices increase. The performance of the condensers can be adversely affected by the build-up of
sediment, scale, corrosion or biological growth inside the tubes. Conventional maintenance
procedures to clean fouled tubes range from chemicals to off-line mechanical cleaners and
brushes, to on-line systems and hydro-blasters. Sometimes, it takes an unconventional,
innovative development in a cleaning system technology to combat ever-changing environmental
challenges. Omaha Public Power District’s North Omaha Station had been drawing water from
the Missouri River for many years with no discernable problems. Recently, they had
experienced a steady rise in summertime backpressures, in the range of 3 to 3.5 in. HgAbs,
resulting in poor condenser performance. Inspections revealed a build-up of calcium carbonate
scale in the tubes. The station has five condenser units with a total of approximately 50,000
stainless steel condenser tubes, all displaying evidence of calcium carbonate build-up. A
recently developed and patented mechanical system for the removal of calcium carbonate scale
without any of the negative drawbacks of chemical alternatives was utilized at the station with
excellent results, proven by subsequent observations. On completion of the process, back
pressure dropped to around 2.0 in. HgAbs and overall performance improved markedly. While
plant personnel strive to determine the cause of calcium carbonate buildup in the condenser,
removing it from the station’s tubes is now a simple procedure and no longer a problem in search
of a solution.
Introduction
Omaha Public Power District’s North Omaha Power Station, shown in Figure 1, has been
operating on the west bank of the Missouri River for many years. The first of its five coal-fired
units has a nameplate output of 75MW. As the region’s population and industrial sectors’ power
requirements grew, additional units were added, with unit 2 at 100 MW, unit 3 at 100 MW, unit
4 at 125 MW, and unit 5 at 200 MW, bringing the plant’s combined capability of its turbine
generators to 600 MW. While the station is primarily base-loaded, one or more units can be
assigned to load-follow, adjusting respective generation up or down to meet demand. The steam
generators, originally designed to burn natural gas or high BTU Kansas coal, now utilize coal
from Wyoming’s Powder River Basin (PRB). The PRB coal has provided the Station with a cost
effective fuel alternative in addition to decreased emission rates.
10937790
219
A key contributor to the efficiency of the power plant steam cycle is the condensation of the
turbine exhaust steam back into a liquid phase that is pumped back to the boiler. The condenser
for each unit, a tube and shell heat exchanger, contains thousands of tubes to perform the
condensing function. Water from the Missouri River is pumped through the tubes in the five
condensers, returning to the river approximately 15 °F higher in temperature after condensing the
turbine exhaust steam. As the steam condenses, the volume occupied by vapor decreases
dramatically to liquid volume, and a vacuum is drawn. The greater the vacuum drawn, the better
the differential pressure across the turbine, increasing the amount of work extracted by the
turbine, thereby resulting in higher cycle efficiency. Poor condenser performance, therefore, can
be identified as one of the single largest efficiency loss factors in power generation. As part of
its maintenance program, in the 1970’s the North Omaha Station replaced all the copper tubes in
their condensers with stainless steel tubes to assure condenser system integrity. It is important
that the river water never breach a condenser tube, allowing contaminants to enter the purified
condensate/steam system. Contaminants could plate out in the boiler tubes and even on the
turbine blades, causing corrosion, build-up, balance and friction problems, any of which would
greatly decrease turbine efficiency and unit availability.
Stainless Steel Tube Cleaning
Stainless steel tubes were first installed for surface condenser applications over 45 years ago.
From Saxon [1], and later Putman [2], we know that these tubes require cleaning. Since the
initial installations, stainless steel condenser tubes have seen many routine cleanings. At the
North Omaha Station, the case had been no different. Routine mechanical and/or hydroblast
cleanings had been performed in order to keep the tube surfaces clean and free from deposits. It
is well known that in addition to inefficient heat transfer if the tubes were not kept clean,
deposits such as iron, manganese, calcium carbonate and bacterial accumulations are prone to
cause crevice corrosion. Therefore, dirty tubes not only increase power generating costs through
lower heat transfer, but also the resulting corrosion failures can lead to a large capital expense to
retube the condenser. While the condensers at the North Omaha Station had been cleaned as
needed in the past to an efficient level of cleanliness, there was a significant loss of vacuum
suggesting an investigation be conducted.
Sources of Performance Degradation
Over the past 4 – 5 years, it was evident the condensers weren’t performing to expectation.
Determining the source of the problem entailed a review of some basic condenser performance
criteria.
Efforts were made to evaluate:
• Turbine backpressure
• Condenser cleanliness factor
• Temperature differential
• Air-inleakage
• Tube leaks
220 10937790
Winter operation, with cooler river temperatures, masked inefficiencies in the condenser.
However, in the spring, as river temperatures began to increase, condenser performance would
suffer significantly as the absolute pressure began to rise noticeably—and only became more
pronounced into the summer/peaking months.
Through outage inspections, conducted by maintenance personnel, the source of the problem,
scale, became evident. What is puzzling however, is that for many years, condenser inspections
typically revealed the tube ID’s to be “gun barrel” clean, referring to a bright, shining surface.
15 – 20 years ago, cleaning consisted of removal of rocks and sticks that had passed through the
traveling screens and onto tube inlets, and hydroblasting silt and loose material from tube
internals. Until recent years, the scale discussed here was not present. Results are explained in
the next section.
The Effect of Calcium Carbonate on Condenser Efficiency
As river water temperatures rose from a wintertime low of 40 ºF to a summertime high of around
84 ºF, the plant observed an increase in condenser absolute, or backpressure, up to 3 to 3.5 in.
HgAbs., well above the expected target of 2 inHgAbs. Efficiency of the turbines was decreasing
as was overall unit heat rate. Another negative effect on the condensate system was seen as the
hotwell temperature increased 10 – 15 ºF with the higher absolute pressure. Generator hydrogen
coolers, using condensate as their cooling meduim no longer had the capacity to adequately cool
the hydrogen gas. This led to periodically limiting the generator output to maintain proper stator
temperatures.
The first and most obvious condition checked was the performance of the condenser air ejectors.
Since air is a good insulator and inhibits heat transfer, a condition called “air-blanketing” in the
condenser tubes must be avoided. The North Omaha Station has both evacuation pumps and
steam jet air ejectors that draw air out of the condensers. These were found to be in good
working order, eliminating an air-blanketing condition as a possible cause. When units are shut
down for planned maintenance outages, before startup, the condenser shell is filled with water
and equipment and tubes are checked for air and river water leaks in a static hydro-test of the
system [3]. During unit 2’s outage cool water used for the condenser hydro apparently
“shocked” the tubes and released sections of scale from the ID of some of the tubes. X-Ray
Fluorescence (XRF) analysis proved the scale to be calcium carbonate, a deposit that, as
previously mentioned had never been a problem from the river water entering the North Omaha
Station. Additional laboratory testing further showed the scale to have a deposit weight per tube
of as much as 184.63 g and a deposit weight density of as much as 24.9150 g/sq.ft. The scale
caused a decrease in the tubes’ ID from .944 in. (clean), down to an average of .925 in., creating
an insulating blanket of scale averaging .020 in. in each of the more than 50,000 tubes in North
Omaha Station’s condensers. The quantified data obtained from X-Ray Fluorescent analysis and
other laboratory testing of the deposit samples is shown in Table I.
10937790
221
Deposit Analysis
Plant: North Omaha Station
Unit: Unit No. 4 Condenser
Tube Info: 1” O.D. x 22 BWG x 30’L
Date of Sample: May 21, 2003
Analysis: XRF (X-Ray Fluorescence)
Elements (%):
Iron .61 Manganese .30
Calcium 92.17 Nickel .002
Magnesium 2.36 Cobalt .002
Titanium .076 Tin 0.00
Potassium .16 Sodium .06
Silicon 2.95 Zinc .004
Aluminum .89 Copper .002
Zicronium .009 Molybdenum .002
Phosphorus .16 Sulfur .23
Loss on Ignition (750 C): 43.17
SAMPLES:
Deposit Wt
Cleaner Deposit Wt. Density
Type Passes (g) (g/sq. ft.)
C3S 1 184.63 24.9150
TABLE I
Deposit Analysis for Unit #4 Condenser
Efficient Removal of Calcium Carbonate Scale
After some trial and error experiments, it was determined that the usual methods of tube cleaning
would prove ineffective with this type of scale and could even prove environmentally harmful in
the case of chemical cleaning. Calcium carbonate (CaCO3) is ranked “3” on the Mohs scale of
hardness [4], about the same as a penny, and is a crystalline compound found in limestone,
marble and shells. Adhering as a stubborn deposit on the inside walls of the condenser tubes, its
removal called for a unique approach. The plant’s management called on the services of Conco
Systems. Having examined the deposit samples sent to them by the North Omaha Station, the
contractor came equipped with a supply of specially manufactured tube cleaners designed to
break the eggshell-like crystalline form of the calcium carbonate deposits. Each tube cleaner has
two rows of four carbide wheels mounted through its circumference that look and act like glass
cutters, scoring the scale as it passes through the tubes. The cleaners were sized to fit inside the
.944 in. ID of the tubes with satisfactory clearance. Figure 2 shows the specially manufactured
and patented tube cleaners.
After opening the water box of unit 5 and exposing the tubes, twenty-five test shots were taken
using a two-gun portable pump system that “shoots” the tube cleaner down the tube at up to 20
ft./sec. with 300 psi water pressure. A lightweight tarp had been hung at the opposite end to
protect the cleaners as well as the water box from damage when the cleaners exited the tube.
Inspection of the tested tubes after shooting showed successful removal of the deposit. Figure 3
shows a close-up photo of the calcium carbonate removed. Figure 4 shows the measurement of a
deposit sample with a caliper; in this image the deposit measures approximately .017 in. thick.
222 10937790
Working around the clock with two four-man crews on 12-hour shifts, the contractor was able to
achieve a high rate of productivity, loading 1,000 tube cleaners at a time in the tubes and
shooting more than 5,000 tubes per shift. Twelve or more passes could be made with each heavy
duty tube cleaner, with recovery at the tarp and cleaning between each pass. Given productivity
rates, each cleaning was completed within the timeframe allocated and no additional outage time
was required. Figure 5 shows a technician “shooting” the condenser tubes. Figure 6 shows
technicians sifting through deposits to recover cleaners for repeat use.
The number and length of tubes vary from condenser to condenser, with 9,240 tubes, length 26’
in unit 1; 8,348 tubes, length 30’ in unit 2 (this unit was not cleaned); also 8,348 tubes, length
30’ in unit 3; 10,900 tubes, length 30’ in unit 4; and 16,485 tubes, length 30’in unit 5. The total
surface area amounts to 415,000 square feet for all five units. Each calcium carbonate-breaking
tube cleaner was followed with a standard tube scraper that performed a final finishing operation.
In all, 18,390 pounds of calcium carbonate deposit were removed from the tubes. Figure 7
shows the piles of deposits removed.
As long as the Missouri River water continues to pose a calcium carbonate deposit build-up
problem, North Omaha Station will perform more frequent and detailed examination of the
condenser tubes, with a preventive maintenance program employing the shooting of standard or
hexagon bladed tube cleaners on an annual or biannual basis. With condenser efficiency
restored, the station reports improved heat rate trends for the affected generating units, with
corresponding optimum power output.
Results
An independent engineering firm was contracted to analyze the results of the cleaning on Units
1, 3, 4 and 5 and report their findings, Unit 2 had been returned to service without cleaning. The
firm commented, “Although we were certain that the tube cleaning had improved the
performance of the Unit 1 and 5 condensers, the detailed operating data provided was too
inconsistent to obtain a reliable measure of the effects.” The following information is extracted
directly from their report [5].
“The North Omaha Station collected Monthly Performance Monitoring Data before and after the
tube cleaning. It was utilized along with condenser design information and the incremental
turbine correction curves to link-up the annual cost benefit of cleaning the condenser tubes. The
apparent condenser heat transfer coefficient was computed the month before and after the
condenser tube cleaning for each Unit.
The values were normalized to a percent tube cleanliness using the industry accepted HEI
algorithm. Then, with the increased tube cleanliness after the scraper cleaning and at an identical
inlet water temperature, design heat load and cooling water flow, the incremental decrease in
turbine exhaust pressure was computed. With linear regressions of the relationship between
turbine backpressure and station heat rate, the percentage improvement in plant heat rate was
determined.
A number of inconsistencies were uncovered with the provided Unit station data. Only Units 3
and 4 had reliable operating condenser performance data that the fuel cost benefits could be
estimated with reasonable certainty.
10937790
223
Despite the relatively low unit cost of coal, the condenser performance improvement for each
unit translated into significant fuel cost savings during subsequent operation as demonstrated in
the Table II below.
Parameter Unit 3 Unit 4

Apparent Cleanliness
34% 37%
(Before Cleaning)
Apparent Cleanliness
93% 90%
(After Cleaning)
Heat Rate Improvement
602.1 395.6
(Btu/KWH)
Annual Savings** $258,400 $212,500
*Data sets taken on April 18, 2003 between 8 and 9 o’clock.

**Assumes a 0.7 Plant Capacity Factor and the average market price for coal.
TABLE II
Fuel Costs Savings
Charts to graphically represent the data for Units 3 and 4 from Table II above are presented in
Figures 8, Improvements in Condenser Performance and Figure 9, Cost Benefits of Cleaning
Operation. In total, they show the improvements associated with cleaning in terms of condenser
performance, overall plant heat rate and annual savings in fuel costs. Note that the annual
savings assumes a 0.7 plant capacity factor and that operation is always above the turbine choke
point with no loss of apparent condenser cleanliness through the year.” The cost of cleaning is
not included in the annual savings calculations.
The annual savings from cleaning Units 3 and 4 were significant. Even though there were
inconsistencies in the data for Units 1 and 5 the amount of calcium carbonate deposit removed
from the tubes was substantial. Additionally, a Unit 2 inspection indicated similar conditions.
Therefore, we estimated that overall savings may exceed $ 1 million a year at the North Omaha
Station.
Conclusions / Recommendations
Changes in local environments can result in heretofore unreported or experienced fouling

problems in power plant condenser tubes. Where this fouling had been previously limited to bio-
forms ranging from algae to mussels, and to mineral deposits such as iron oxide, several plants
have been reporting the occurrence of extremely hard scaled calcium carbonate. While the cause
of scale buildup – just in recent years – has not yet been determined, steps can be taken to relieve
the problem.
Recommendations are:
1. Monitor and trend condenser performance on a routine/frequent basis.

2. Examine tube ID’s during outage periods and, if present, identify deposits. If calcium
carbonate scale is present, try using standard heavy duty tube shooting system tube
cleaners. This may alleviate the problem if the scale layer is minimal.
224 10937790
3. If the scale layer has exceeded the ability of standard cleaners to remove it, contact a
service provider meeting the certification requirements as noted above for the removal of
calcium carbonate scale.
4. Upon successful removal of the scale, perform frequent monitoring and preventive
maintenance activities that include preemptive detection and elimination.
Acknowledgments
The authors wish to acknowledge and thank the Omaha Public Power District and Conco
Systems, Inc. for sharing this case with industry.
References
[1] Saxon, G. E., Stainless Steel Tube Cleaning, Conco Bullets, Vol. 3, Issue No. 7, 1988.
pp 1-2.
[2] Putman, R. E., Steam Surface Condensers: Basic Principles, Performance Monitoring,
and Maintenance, 2001 ASME Press, New York, NY, pp 173-174.
[3] Woodruff, Lammers, Lammers – Steam Plant Operation, 7th edition
[4] Memphis Archeological and Geological Society - Mineral Hardness Table ©1999
[5] Report on Fuel Cost Savings Resulting From Condenser Tube Cleaning North Omaha
Station, Burns, J.M and Burns, D.C., Burns Engineering Services Inc., Topsfield, MA,
November 2003
Figure 1
North Omaha Power Station
10937790
225
Figure 2
Specially Manufactured and Patented Tube Cleaners
Figure 3
Close-up of Calcium Carbonate Removed
Figure 4
Measuring Deposit Sample with Caliper
226 10937790
Figure 5
Technician Shooting Tubes
Figure 6
Technicians Sifting Deposits For Cleaners
Figure 7
Piles of Calcium Carbonate Deposits
10937790
227
Improvement in Condenser Performance
100
93
90
Apparent Cleanliness (%)

90
80
70
60
Before Cleaning
50
After Cleaning
37
40
34
30
20
10
0
3 4
Unit
Figure 8
Improvements in Condenser Performance
Benefits of Cleaning Operation
700 300,000
600
Reduced Heat Rate (Btu/Kwh)
250,000
Savings in US Dollars ($)
500
200,000
400
Heat Rate
150,000
Annual Savings
300
100,000
200
50,000
100
0 0
3 4
Unit
Figure 9
Cost Benefits of Cleaning Operation
228 10937790
4
SESSION 4: FOULING CONTROL
10937790
229
10937790
Condenser Biofouling at Entergy’s Gerald Andrus Power Plant
Diagnosis and Control Strategy
By
Darrell R.Coker
Applied Specialties, Inc.
Avon Lake, Ohio
Paula Weaver
Entergy
Gerald Andrus Plant
Greenville, Mississippi
ABSTRACT
A large gas/oil fired power plant in Greenville, Mississippi began to experience severe
main surface condenser fouling causing a peak megawatt loss of over 90 megawatts in
the summer of 2003. During 2003 the plant mechanically cleaned the condenser on four
occasions, only to experience fouling reoccurrence and loss of performance. This paper
discusses the historical trends of the fouling and the testing that eventually led to a
successful solution to the problem. The condenser is a two-pass design composed of 316
stainless steel tubing. Plant cooling is essentially raw Mississippi River water.
Onsite side-stream monitoring and analytical testing was initiated in the summer of
2003 for root cause determination and potential solution development. Testing included
a side stream heat exchanger, water sampling, biological testing, deposit analyses, and
microbiological DNA analyses. The plant had initiated a dispersant program based on
vendor recommendations with no success.
Significant growth was noted on the test heat exchanger and tubing within a few days
of testing with significant fouling noted within two weeks. This fouling rate was
consistent with the actual plant condenser cleanliness trends. Deposit analysis of the
cleaning debris revealed over 60% iron and manganese in the sample. DNA results on
the foulant also documented high iron and manganese fixing bacterial counts. DNA and
elemental analysis was also performed on the test heat exchanger foulant and found to be
consistent with the actual condenser deposition.
Several non-oxidizing biocides were tested in the laboratory and one was selected for
further testing based on its’ superior effectiveness on the iron and manganese bacteria.
The biocide was tested on the side-stream monitor and found to be extremely effective in
the actual plant system. Environmental approval was requested and approved for full
scale plant trial in 2004.
Due to the high cost of non-oxidizing biocides, a method of treating the condenser off-
line was designed to reduce the overall program costs. The plant is a peaking unit that
10937790
231
comes down to minimum load at night. A system was designed and implemented to
inject the biocide to a single loop of the condenser (half of the condenser) in a closed
loop type style. An automated injection installed allowing the operators to isolate the
butterfly valves on one loop of the condenser, push a button to inject and recirculate the
biocide treated water. Actual plant treatment was implemented in the spring of 2004.
During 2004 the plant operated with no loss of condenser performance during the
entire summer and no mechanical cleaning was required. No fouling has occurred in
2005 year-to-date.
INTRODUCTION
Entergy’s Gerald Andrus Steam Electric Station is located just south of Greenville,
Mississippi on the banks of the Mississippi River. The plant was began operation in
1974 and can produce 781 megawatts of electricity from either natural gas or oil. It is a
supercritical unit using raw Mississippi River water to supply the large once through
cooling condensers. Circulating water flow is approximately 200,000 gallons per minute
and condenser metallurgy is 316 stainless steel.
Condenser design is two-pass composed of an inner and outer loop with a delta T of
approximately 10 degrees F on each pass. Condenser integrity has historically been
excellent with few tube failures experienced during the 30 years of operation. Fouling
had also not been a problem until year 2000 when the plant began to experience some
light to moderate condenser fouling. Plant personnel credited this new fouling to river
siltation and mechanically cleaned periodically to maintain performance. Between 2000
and 2003 the fouling problem became so severe the unit was forced to de-rate by 90+
megawatts. Contract mechanical cleaning was performed during the summer of 2003 on
four occasions, only to experience re-fouling within two weeks. Cost for mechanical
cleaning during 2003 was over $160,000 and required a 3-4 day outage to perform the
work.
Gerald Andrus’ biofouling program had historically been 1 hour per day chlorination
per pass for a total of two hours per day. BCDMH (a chlorine/bromine oxidizing
biocide) was added to achieve a free chlorine residual of approximately 0.15 FAC. Their
water treatment vendor recommended the plant make some treatment changes during
2002 which were continued into 2003. The bromination program was converted to six (10
minute) injections every four hours versus the historical one hour daily continuous
treatment. This change was in response to high manganese levels found in some deposit
samples taken during the year. The focus was on preventing the oxidation of manganese
from the oxidizing biocide and suspending any solid material in the water with
dispersants. None of these changes were successful in 2003 and the fouling continued to
plague the plant in 2003.
232 10937790
ONSITE PROBLEM IDENTIFICATION AND ANALYTICAL TESTING
During the early summer of 2004, Applied Specialties became involved in solving
Gerald Andrus’ condenser fouling problems. A side-stream test apparatus was installed
at the condenser that included a section of clear poly tubing leading from a valve on the
condenser outlet to a test heat exchanger. This test heat exchanger was heated and
contained a 316 SS tube with approximately the same flow rate as the condenser
operating conditions.
After only a few days of operating this system the test heat exchanger and clear poly
tubing were noticeably fouled. Within two weeks the tubes were so severely fouled that
one could no longer see through the poly tubing. Samples of the tubing were sealed and
submitted for elemental and biological DNA analysis along with samples of past cleaning
debris. Our intent was to verify that the fouling occurring in the test system was
consistent with the actual foulant being removed from the condenser during mechanical
cleaning. Water samples from the Mississippi River were also analyzed to determine the
bacterial and elemental composition during the test period.
The foulant was identified as 48.6% manganese and 12.6% iron. The remainder of the
elemental analysis was consistent with typical river silt composition of silica, calcium,
aluminum, and magnesium. DNA analysis indicated very high percentage of iron and
manganese fixing bacteria in the condenser cleaning sample and the test loop. The
results were very consistent in both samples lending validity that our side-stream test was
simulating the actual plant condenser conditions.
10937790
233
Deposit Analysis
Results
APPLIED SPECIALTIES, INC.
Client: Entergy - Andrew Plant, Greenville, MS Lab Number: 020812-B
Source of Sample: Deposit Date Sampled: 8/6/2003
Date Received: 8/12/2003 Date Reported: 8/16/2003
Deposit Analysis Results (as received basis)

% Moisture @ 105 Degrees C 5.8%
% Fix Solids @ 650 Degrees C 75.7%
% Volatile @ 650 Degrees C 18.5%
% Fixed Solids @ 900 Degrees C Possible Composition

Manganese as MnO2 48.6%
Aluminum as Al2O3 16.0%
Iron as Fe2O3 12.6%
Calcium as CaO 6.4%
Magnesium as MgO 3.6%
Potassium as K2O 2.2%
Silicon as SiO2 1.9%
Other Oxides & Neutral Salts 2.0%
234 10937790
DNA RESULTS
Project: Bacterial DNA Analysis of Deposit

Entergy – Andrew Plant, Greenville, MS
Sample Description: Bacterial DNA Analysis of Deposit
Date Sampled: 8/22/03
Sample ID: #020812-B
RESULTS:
Bacterial isolates identified using Genus-specific DNA hybridization probes
Total Calculated Bacterial Count Based on DNA/probe complex ∼2.55e+3 cfu/mg (As-Is Wt.)
Total Calculated Aerobic Bacterial Count ∼1.27e+3 cfu/mg (As-Is Wt.)
Bacteria-Slime Formers Spore Formers (8.7%)
Clostridium sp. (Anaerobic – Non H2S Producing) – 7.1% [∼1.81e+2 cfu/mg (As-Is Wt.)]
Bacteria-Slime Formers Nonspore Formers (49.7%)
Enterococcus sp. (Aerobic) – 0.6% [∼1.60e+1 cfu/mg (As-Is Wt.)]
Proteus sp. (Aerobic) – 1.3% [∼3.3e+1 cfu/mg (As-Is Wt.)]
Acinctobacter sp. (Aerobic-denitrifying) – 3.2% [∼8.2e+1 cfu/mg (As-Is Wt.)]
Serratia sp. (Aerobic) – 5.2% [∼1.32e+2 cfu/mg (As-Is Wt.)]
Alcaligenes sp. (Aerobic – denitrifying) – 6.5% [∼1.65e+2 cfu/mg (As-Is Wt.)]
Achromobacter sp. (Aerobic – denitrifying) – 5.2% [∼1.32e+2 cfu/mg (As-Is Wt.)]
Pseudomonas acruginosa (Aerobic – denitrifying) – 3.2% [∼8.20e+1 cfu/mg (As-Is Wt.)]
Pseudomonas cepecia (Aerobic) – 2.6% [∼6.60e+1 cfu/mg (As-Is Wt.)]
Pseudomonas fluorescens (Aerobic – denitrifying) – 4.5% [∼1.15e+2 cfu/mg (As-Is Wt.)]
Pseudomonas paucemoblia (Aerobic) – 3.9% [∼9.90e+1 cfu/mg (As-Is Wt.)]
Pseudomonas mendocina (Aerobic) – 5.8% [∼1.48e+2 cfu/mg (As-Is Wt.)]
E. coli (Aerobic) – 7.7% [∼1.98e+2 cfu/mg (As-Is Wt.)]
Bacteria-Corrosive Iron Depositing (19.4%)
*Leptothrix sp. (Aerobic) – 5.2% [∼1.32e+2 cfu/mg (As-Is Wt.)]
*Sphaerotilus sp. (Aerobic) – 5.8% [∼1.48e+2 cfu/mg (As-Is Wt.)]
Gallionellis sp. (Aerobic) –8.4% [∼2.14e+2 cfu/mg (As-Is Wt.)]
Bacteria-Corrosive Sulfate Reducers (18.1%)
Desulfovibrio sp. (Anaerobic – H2S Producing) – 5.8% [∼1.48e+2 cfu/mg (As-Is Wt.)]
Clostridium sp. (H2S Producing - Anaerobic) – 12.3% [∼3.13e+2 cfu/mg (As-Is Wt.)]
Bacteria-Corrosive Sulfuric Acid Producers (5.8%)
Thiobacillus sp. (Aerobic – some strains denitrify and most strains produce sulfur) –5.8% [∼1.48e+2 cfu/mg (As-Is Wt.)]
* Iron Bacteria Leptothrix and Sphaerotilus Deposit Manganese
10937790
235
The side stream test system was altered to include two chemical injection points and
pumps to test various treatment options. We decided to disprove dispersant technology
as a possible solution and test the biocide that had been selected using lab tests.
Sidestream chemical injection testing began in the late summer of 2003. Dispersant was
fed at five times the recommended dosage and biocide was fed at a continuous 5 ppm
dosage. Within a week it was obvious the dispersant was ineffective in controlling the
fouling even at the highly elevated dose. The hose section and test heat exchanger
located downstream from the biocide injection remained very clean, clear, and un-fouled.
Further testing was performed using the biocide injection in an effort to optimize the
dosage and frequency. We injected 1.0 ppm biocide continuously and continued to see
no fouling on the hoses or test heat exchanger. Timers were installed and testing
continued successfully using an intermittent approach (30 minutes on and 30 minutes off
scenario). As fall approached and Mississippi River water temperatures decreased, the
fouling also decreased, resulting in less than optimum testing conditions. Testing was
discontinued in October of 2004 and the results were submitted to management for
review.
SOLUTION IMPLEMENTATION
Due to the severity and costliness of the condenser fouling, plant management decided
to pursue the non-oxidizing biocide option for implementation in the spring of 2004.
Several meetings and brainstorming sessions were held to determine our best path
forward. During one of these meetings plant operations indicated they could safely
isolate one of the condenser loops at night and continue to operate safely at minimum
load. Given this information the decision was made to submit an application to the EPA
for approval of the biocide. We also began to design a feed system that would be simple
to use and relatively failsafe for the operations personnel.
Environmental approval was obtained and equipment was installed in the early spring
of 2004. The equipment consisted of PVC piping connecting the drain lines for the two
condenser loops. Selenoid and manual valves were installed at the various locations and
a PLC controlled control box was fabricated. A single pump was installed to provide
circulation within the boxes and also to inject the non-oxidizing biocide into the
condenser inlet area. Pump suction was aligned to the tote bin during chemical feed and
would subsequently be switched to recirculation after the injection. The timer on the PLC
was initially setup to inject the biocide for 15 seconds after which the chemical tote
would be isolated and recirculation between the water boxes would begin. Another PLC
timer controlled the recirculation timer and operated the selenoid valves within the
system. Using this design the operators simply closed the appropriate butterfly valves
leading into and out of either “A” or “B” loops of the condenser and pushed the chemical
injection button. He could then come back 4-6 hours later, secure the system, and re-
open the condenser butterfly isolation valves and resume normal operation. This process
was performed nightly while alternating loops every other night.
236 10937790
Dispersant injection was eliminated due to the lack of performance at 5 times the
recommended dosage under test conditions. The bromine program was reinstated at one
hour continuous daily per condenser loop. A mass balance of manganese and iron using
the river water makeup levels (less than 30 ppb) and flow was performed compared to the
amount of iron and manganese removed from the condenser in 2003 using mechanical
cleaning. This mass balance clearly proved the oxidizing biocide could not have been a
major cause of the deposits (by oxidation) and that the fouling process was occurring 24
hours a day versus one hour daily during bromination.
RESULTS
During 2004 the condenser operated per design throughout the entire summer period.
No de-rates occurred due to condenser backpressure and no mechanical cleaning was
required. Boroscope and visual inspections revealed the condenser was extremely clean
and free of deposits. During the initial boroscope inspection following summer
operation, corporate technical personnel performed an inspection and thought they had
arrived too late to do a pre-cleaning (mechanical) inspection as they had done in the past.
Because of the extremely clean condition of the condenser, they assumed the condenser
had already been mechanically cleaned (and labeled their taped report as post-clean).
They also noted in the written inspection report to plant management that the condenser
was cleaner than they had ever experienced in their history of inspections at this site.
Results for 2005 year-to-date are consistent with the results for 2004. No fouling has
been noted through August, 2005.
10937790
237
10937790
ON-LINE AUTOMATIC TUBE CLEANING SYSTEM
AND
ON-LINE SELF FLUSHING DEBRIS FILTER
Kaveh Someah
Brackett Green USA, Inc.
Houston, Texas
(281) 480-7955
ksomeah@bgusa.com
ABSTRACT
This paper describes various types of On-Line Automatic Tube Cleaning Systems
(ATCS) and Self Flushing Debris Filters (DF) in Power plane application.
Macro and micro-fouling as well as scaling and deposit inside the condenser tubes
significantly reduces the condenser’s efficiency and cleanliness factor while impacting
plant performance. According to various studies, fouling and scaling inside the tube
contributes up to 50 percent of the total condenser tube heat transfer resistance.
The use of On-Line Self-Flushing Debris Filters as well as the On-Line Automatic Tube
Cleaning Systems is an effective way to continuously and permantly eliminate the macro
and micro-fouling ad well as scale and deposits from the condenser tube.
INTRODUCTION
Plant availability, increased output, reduced fuel consumption, reduced operating and
maintenance cost, protecting the water and air quality have become an important aspect
of all power plants worldwide. The goals include improved thermal performance,
increase plant output, and reduce environmental impact, while increasing plant
profitability. Each plant, depending on the source of the cooling water used faces a
different challenge.
Considering the age of most generating plants here in the U.S., the cooling water system
including the intake systems whether open loop (river, sea) or closed loop (cooling
tower, pond) experience problems. Condensers are either old and under sized due to
loss of tube surface (plugged) or have been retubed with alternative material that is more
susceptible to micro-fouling and scaling. The traveling screens at the intake require
continuous maintenance and allow debris carry over through the screens. In some
plants where cooling towers are used the deterioration of the cooling tower fill or the
structure is causing the tube blockage. Furthermore, plants are more restricted in use of
traditional chemical treatment. In many areas of the country there has been change in
the aquatic marine life, forcing the plants to battle unusual debris presence within the
cooling water intake system. Although the use of the dual/central flow or the disc type
intake traveling screen can reduce the potential for the debris carry over, however, many
marine life type debris as well as grass passes thorough the screen opening and
10937790
239
attached/grow inside the cooling water intake tunnel wall. They are later released and
deposited inside the condenser tubes.
The limitation on chemical treatment, regulation on zero chemical discharge, coupled

with degradation in quality of cooling water source has resulted in increased condenser
fouling scaling.
In 1981, EPRI surveyed 415 fossil fueled Power Plants in the U.S. The survey
determined the bio-fouling as the dominant cause of availability loss. According to
various studies and reports published, the water side fouling of the condenser tubes
contributes up to 50 percent of the total resistance to heat transfer. This resistance to
the heat transfer is caused by the laminar and stagnant layer of water next to the tube
wall as well as the build up of deposit, or scale. The deposits often consist of slim,
calcium carbonate, magnesia dioxide, corrosion products, silt, petroleum products,
bacteria matrix, etc. In recent years many plants have retubed the condensers with
material such as stainless steel, or titanium that do not contain copper. Such tube
materials are highly susceptible to biological fouling.
MICRO-FOULING AND SCALING
The accumulation of fouling and scale will have negative impact on the plant heat rate as
well as tube life. Scaling and fouling will eventually cause under deposit corrosion over
the tube surface area. The tube corrosion and the MIC attack will reduce the tube life
and cause frequent outage due to tube leaks. Case studies have also demonstrated the
additional fuel consumption and increase in CO2 and NOX emission for plants that
operate with lower condenser cleanliness factor and higher turbine back pressure.
The off line cleaning of the condenser requires manual “shooting” of the tubes with high
pressure water/air, brush or scrapers. The process requires periodical plant shutdown
and loss of production. In some applications the manual cleaning could only be partially
effective and the fouling and scaling could form rapidly after each clean cycle.
Furthermore the off line cleaning may not be very effective since it is not eliminating the
major heat resistance that is caused by the stagnant water flow next to the tube surface.
On-Line Automatic Tube Cleaning Systems (ATCS) have been an effective method in
eliminating scaling, fouling, while increasing the tube life and improving the plant heat
rate.
One type of on-line cleaning system consists of a set of polypropylene baskets (cages)
that are epoxyed at the end of each tube. A single brush with nylon bristles is inserted
inside each basket.
240 10937790
Figure 1 shows the installation in a typical condenser. During normal condenser
operation, the brushes are held in the down stream cage without interfering with the
condenser water flow. The pressure drop caused by the brush and the cages are
negligible.
Figure 1
Typical Installation For The Brush And Basket Type On-Line Automatic Tube Cleaning
System
On-line cleaning of the condenser can be accomplished periodically by reversing the

cooling water flow through the condenser. Each time the condenser flow is reversed
(back flushed); the brushes travel the full length of the tubes and positively brush each
tube, removing any scale or deposits. Figure 2 shows typical piping arrangement and
condenser back flushing employed by some plants.
10937790
241
Figure 2
Typical Flow Reversal Mechanism
Another effective system which eliminates condenser micro fouling and scaling is the
circulation of slightly oversized elastomeric (rubber) balls. Uniquely designed slightly
oversized elastomer balls (Figure 3) are periodically or continuously injected into the
condenser cooling water inlet.
Figure 3
Typical Elastomeric Balls
242 10937790
The balls are passed through the condenser tubes by the cooling water flow. A typical
system is illustrated in Figure 4.
Figure 4
Ball Type Automatic Tube Cleaning System
The balls are designed and injected in such a way to provide a uniform distribution inside
the water box. Since the diameters of the sponge balls are larger than the inside
diameter of the tubes, the condenser tubes are automatically kept clean by the
scrubbing and wiping action of the balls which prevents the deposit of micro-foul ants on
the inside tube surface. The balls are removed by means of a strainer installed in the
outlet condenser pipe Figure 5 shows a typical strainers installed at the cooling water
outlet pipe. From there they are returned and re-injected into the condenser inlet line.
Figure 5
A Typical Ball collection strainer installed at the cooling water outlet
10937790
243
The major components of the system are the ball strainer (Figure 6) and the ball
recirculation control system. Ball and water separation is accomplished by two half
elliptical screens housed in as spool piece and held in a rigid position in relation to each
other within full diameter of the pipe section. Each screen is mounted on a shaft and
can be turned to the backwash position automatically (or manually) by an exterior
motorized actuator. The two shafts and the motorized actuators operate independently
of each other. The two screens are situated in a manner in which each screen makes a
separate hopper with each half of the pipes cross sectional area. Side-blank off plates
are attached to pipe directly. These plates along with the two half-sized elliptical
screens make up the hopper. In addition, specially designed baffle plates are installed
directly on the pipe section and the screen surface, the area where the elliptical section
of the pipe faces against the curvature of the pipe. The baffles not only eliminate the
potential for the presence of dead spots where the balls could collect and not be
recovered. The baffles create a vortex and turbulent flow that forces the balls to be lifted
from the screen surfaces and forces them toward the ball extraction ports.
Figure 6
Typical Strainer Section In The Closed (Operating) And Open (Back flushing) Condition
After leaving the strainer sections, the balls pass through a special recirculation pump,
which adds sufficient pressure to overcome the condenser and recirculation piping
pressure drop. The balls are pumped into the ball collector. Figure 7 presents a typical
ball recirculation pump and the ball collector chamber.
Figure 7
Ball Recirculation Pump Skid And Control System As Well As Control Panel
244 10937790
The balls are then distributed into the cooling water supply lines by means of the
injection nozzles. Figure 8 presents a typical ball injection nozzles and the ball injection
into the two cooling water supply inlet. The ball count is continually monitored and
displayed.
Figure 8
Ball Injection Nozzle Into the Two Cooling Water Inlet
The ball collector is used to automatically put the balls into or out of circulation. The
balls are also kept in this chamber during the strainer backwashing. The collector also
serves as a means to replace worn out balls or change balls when needed.
The screen backwashing sequence automatically starts when the differential pressure
across the screen reaches a preset level. Balls are automatically collected inside the
collector prior to any backwashing. This increase in the differential pressure is caused
by debris and particles present in the circulating cooling water. The differential pressure
across the screens is continually monitored and displayed. Balls are automatically
collected inside the collector prior to any backwashing.
All functions and operations of the ball system are controlled and displayed by means of
a single central control panel using a series of PLCs. It controls the operation of each
screen, the pump and the collector. The control system offers total flexibility to operate
the ball cleaning system on an automatic continuous or automatic batch intermittent
cycle. The system can also be manually operated at the control panel.
MACRO-FOULING
In general, he condenser macro-fouling consists of accumulation of solid debris such as

fish, clam, muscle, barnacle, oyster shell, plastic, wood, pine cone, leaves and
occasional jelly fish as well as fibrous grass, seaweed, algae, etc., which clogs tube
sheets and the tubes, reducing the water flow though the condenser. To elevate the
problem the plants often need to partially or fully reduce the load and perform frequent
manual “picking” of debris from each half or the entire condenser. At locations where
the water is subject to severe storms, typhoons, hurricanes, or any tidal wave currents
as well the churning of the water due to passing ships or the presence of storm drain or
10937790
245
water treatment plants upstream to the plant, the accumulation of debris could be
significant.
In addition to adverse effects on the plant’s thermal rate, the accumulation of debris will
cause erosion of the tubes and the tube sheets. The debris is often partially lodged
inside the tubes, causing flow impediment and local erosion of the tube surface and the
tube leaks.
Figure 9
246 10937790
Figure 9
Condenser Tubes and Tube Sheet Fouled with Debris and Some of the Debris is
Removed.
Figure 9 shows the condenser tube and tube sheet at one utility. The plant operates two
(2) identical 350 MWe coal fired units. The cooling water for the plant is drawn from a
river through a common tunnel. To keep the condensers clean, the plant has to routinely
and as often as once every two weeks reduce the load partially and remove the debris
from the water boxes manually, one half water box at a time. The plant estimated that
they remove up to four (4) 55 gallon drums from both units every two weeks and
average power loss of 18 MW/units due to debris accumulation. In 2003 the plant
installed an On-Line Debris Filter on Unit 1. Figure 10 shows the condition of the tubes
and the tube sheet when the unit was inspected after 6 months of continuous operation.
During the six (6) month period, Unit 1 remained on line continuously without any drop in
capacity or increased back pressure. The debris filter can be partially seen at the
condenser inlet nozzle. Based on this result the plant has now completed the installation
of the debris filter on Unit 2.
Figure 10
Clean Condenser Tubes and Tube Sheet After Debris Filter Installation
10937790
247
The debris filter at this or other plants are in the form of a short spool piece that is
installed at the condenser inlet water nozzle as shown in Figure 11
Figure 11
Typical Debris Filter Insallation
The spool piece replaces a section of the existing cooling water pipe. In some
installations the debris filter has been installed in the form of an elbow that has replaced
the present elbow connection to the water box. Figure 12 shows the installation of such
debris filter in a straight or elbow section. As shown in this figure, the built-in-elbow type
replaces an exisiting elbow, which may exsit. Such installation will require no piping
change or piping modification
248 10937790
Straight Section Installation
Before Debris Filter Installation
After Debris Filter Installation
Figure 12
Debris Filter Installation In Various Orientations
10937790
249
The debris filters principal of operation is based on pressure relieved backwashing
concept.
The filter consists of a housing structure that contains the debris collection screen and
the debris removing suction port. Figure 5 represents a typical outline drawing of such
filter. The filter screen (1) is installed in the form of a perforated, wire mesh, wedge wire,
or metal grid, inside the housing. The screen gap is selected between 5 to 10 mm,
depending on the application. The incoming debris is collected against the filter screen.
The filter also includes a debris evacuation rotor (2), a debris suction nozzle (3), a rotor
drive motor (4), a debris discharge pipe (5), a debris discharge valve, and a differential
pressure measuring system (7).
Figure 5
A Typical Debris Filter Structure
The accumulation of debris on the filter screen creates a differential pressure across the
filter screen higher than the “dirty screen” set point. Once the differential pressure
reaches this set point, the cleaning cycle sequence will be initiated. The actuated
discharge valve will be open and the discharge rotor will begin to turn. The differential
pressure at the portals creates a vacuum (flow reversal), which will remove the debris
from the screen. The discharge rotor continues to clean the entire screen surface until
the monitoring system indicates a clean condition.
The debris filter is designed to operate with adding negligible additional pressure drop to
the existing condenser unit and will use a small quantity of the cooling water during the
back-flushing. The filters are proven to require very little or no annual maintenance.
CONCULSION
Condenser micro and macro-fouling as well as scaling and deposit inside the condenser
tube adversely impacts the plant thermal performance and availability.
The use of On-Line Automatic Debris Filters and On-Line Automatic Tube cleaning
System has proven to be an effective way to improve plant performance, increase output
and availability while reducing the operating and maintenance cost.
250 10937790
THE PRACTICAL APPLICATION AND INNOVATION OF CLEANING TECHNOLOGY
FOR CONDENSERS
Andy Howell George Saxon, Jr.

Xcel Energy Conco Systems, Inc.
4653 Table Mountain Drive 530 Jones Street
Golden, CO 80403 USA Verona, PA 15147 USA
andy.howell@xcelenergy.com gsaxonjr@concosystems.net
Abstract
Maintaining clean condenser tubes is of vital importance for reliable, efficient power generation.
Two major problems result from substances on the interior of tube surfaces: (1) Loss of heat
transfer, and (2) Under-deposit corrosion. Internal tube fouling is nearly always detrimental to
heat transfer, thus reducing the efficiency of steam condensing, resulting in a lower vacuum
(higher pressure) and less efficient steam turbine operation. Under-deposit corrosion of tube
material can result in through-wall leaks, permitting ingress of cooling water into high purity
steam condensate. Because poor vacuum conditions in the condenser can reduce electric
generating capacity, and contaminants in steam condensate can cause significant damage to
boiler tubing and steam turbine materials, measures to prevent fouling or remove foulants must
be applied. Early identification of fouling characteristics and a fundamental knowledge of
cleaning system capabilities are essential in determining the most effective cleaning technology.
Unique circumstances may require innovative solutions. The common causes of condenser tube
fouling and the practical application and innovation of cleaning technology are discussed. State-
of-the-Art methods for off-line mechanical tube cleaning are presented, and their effects on tube
condition examined.
Introduction
Maintaining clean condenser tubes is of vital importance to reliable, efficient power plant
operation. Ideally, fouling would be avoided entirely; however, this is rarely the case.
Condensers may become lightly fouled with soft organic deposits, or can be severely scaled with
hardened minerals. The probability of success in cleaning the condenser is dependent on the
selection of the appropriate cleaning technology under the specific fouling conditions. Early
identification of fouling characteristics and a fundamental knowledge of cleaning system
capabilities are essential in determining the most effective cleaning technology, as well as the
frequency of cleaning required.
Two major problems result from substances on the interior tube surfaces: (1) Loss of heat
transfer, and (2) Under-deposit corrosion. Internal tube fouling is nearly always detrimental to
heat transfer, thus reducing the efficiency of steam condensing, resulting in a lower vacuum
10937790
251
(higher pressure) and less efficient steam turbine operation. In severe cases, poor vacuum
conditions in the condenser can reduce electric-generating capacity by more than fifty percent.
Under-deposit corrosion of tube metal can result in through-wall leaks, permitting ingress of
cooling water into high-purity steam condensate. Contaminants in cooling water (which may be
concentrated significantly in a recirculating cooling tower) can cause major damage to boiler
tubing and steam turbine materials, potentially resulting in huge economic penalties in unit
outages and equipment repair and/or replacement. Measures to remove foulants from condenser
tubes can be a very minor price to pay in terms of the overall picture.
Options for cleaning foulants off-line include chemical (acid or chelate dissolution) and
mechanical means (metal or plastic scrapers, brushes, high pressure water, etc.). On-line
preventive measures also exist, including chemical treatments (scale and corrosion inhibition,
dispersants, biocides) and continuous recirculation of mechanical cleaners of a variety of types.
A close look will be taken at the use of metal scrapers for off-line cleaning, as this has been a
reliable and effective technology for several decades, with scraper designs continually evolving
to meet the needs of various fouling situations. Metal scrapers have distinct advantages over
other tube cleaning methods in certain scenarios. The most effectively designed scrapers are
sized to closely match tube interior dimensions in order to clean near to the base metal. Scrapers
are propelled through the tubing at high velocity (20 ft/sec) by pressurized water. Occasionally,
when difficult deposits and corrosion products are encountered, repeated passes of the scrapers
may be used to insure optimal cleaning. Innovation of new tube cleaner designs may also be
required to enable effective cleaning of unique deposit scenarios, or when existing technology is
inadequate. Since condenser fouling can have such a dramatic impact on cost-effective power
plant operation, deposition and corrosion should be carefully monitored, and the condenser
cleaned at optimal intervals.
Condenser Tube Fouling
The principal modes and types of fouling on the interior of condenser tubes are:
• Deposition or particulate
• Scaling or crystallization
• Microbiological
• Debris or macrofouling
• Corrosion and corrosion products
In most cases, fouling is not due to a single mechanism; however, in many situations one
mechanism will be dominant. Fouling tends to increase over time, with the rate of increase being
very site-specific. Recognizing this, the Tubular Exchanger Manufacturers Association [1] has
recommended that designers of heat exchangers include an allowable fouling resistance in their
calculations, in order that some fouling can be tolerated before cleaning becomes necessary.
Heat Exchange Institute standards [2] also include some fouling in their requirements for steam
surface condensers. Although these allowances tend to mask losses of efficiency, fouling still
may have an impact on under-deposit corrosion; most corrosion problems on the interior of
condenser tubes result from tube fouling. Thus, determining when to clean often requires
striking a balance between tolerable efficiency loss and providing protection from corrosion. A
model of under-deposit corrosion [3] is shown in Fig. 1.
252 10937790
Deposition or Particulate Fouling
Deposited material is fine-particulate debris that settles on the tube surface due to gravity under
low-flow conditions. This debris may be natural sediment (e.g. river silt), biogrowth (e.g. algae),
coal dust from plant operations transported to source water, or crystalline solids precipitated as
particulates in a cooling tower basin (e.g. calcium carbonate, silicates, phosphates, sulfates, etc.).
Deposits may form inconsistently over a tube surface, primarily occurring at the bottom of the
tube due to gravity settling.
Normal recommended flows through condenser tubes (minimum seven feet per second) should
be adequate to prevent the settling of particulates; however, a condenser with an average flow
representing 7 fps may have areas or specific tubes with much lower flow rates. For example,
inadequate circulating water pumping may not entirely fill an inlet waterbox, so that intermittent
and partial flow can occur in upper tubing. Partial blockage will also cause reduced flow
through tubes, which can result in considerably increased particulate settling and accumulation
downstream of the blockage.
The most serious problem resulting from deposition in condenser tubes is under-deposit
corrosion. This can be a serious problem even when heat transfer efficiency is only a minor
concern, because the amount of tube surface containing deposits that stimulate under-deposit
corrosion can be relatively small. Corrosion occurs primarily because of the electrochemical
potential differential between the tube surface beneath the deposit and the adjacent general
(clean) tube surface. Additionally, corrosion is often enhanced by the action of bacteria within
the deposit, which can produce corrosive metabolic constituents. Because of this phenomenon it
is necessary to keep tubes as clean as is practical.
Scaling or Crystallization Fouling
Scale occurs when the saturation point of dissolved constituents in the cooling water is exceeded,
and precipitation occurs directly on tube surfaces. Some common cooling water constituents
form scales that precipitate preferentially under higher temperature conditions (calcium
carbonate, calcium phosphate); therefore, formation of scale on warm tube surfaces is promoted.
Such scaling is also more likely to occur on hotter tube surfaces (towards the condenser outlet).
When condenser tubes are scaled it is usually a widespread occurrence, and very detrimental to
heat transfer. Even a thin layer of scale can cause a major reduction in heat transfer; for instance,
manganese, silica and calcium carbonate scales have been reported to cause losses in output
capacity of 20 to 25 megawatts [4]. Scale can also provide sites for local under-scale corrosion
(Fig. 2). Inconsistently-present scale, with areas of scale-free surface due to either loss of
existing scale or inconsistent formation, is particularly detrimental; erosion-corrosion can result
in brass tubes due to local flow turbulence [5] (Fig. 3), and the likelihood of damage from
differential concentration corrosion cells may also be enhanced. The most comprehensive
removal of scale is generally required.
Microbiological Fouling
Bacteria present in natural waters will commonly form a thin biofilm on condenser tube surfaces.
Recirculating cooling water (cooling towers) is typically treated with biocidal chemicals, such as
10937790
253
bleach, chlorine gas or organic biocides. These may be fed continuously or intermittently (e.g.,
two hours daily), and are very effective in limiting the growth of microorganism populations.
However, once a biofilm gains a foothold, biocides may not be effective for film removal.
Biofilm development is inversely related to flow velocity; attachment of bacteria to the substrate
is more difficult at higher velocities. Conversely, low velocities (similar to those that may
support fine particulate deposition) are conducive to bio-growth. Biofilms contain a
considerable amount of entrapped water, and when dry, may appear to be of minimal
significance; however, water is a relatively good insulator of heat, so even thin biofilms can have
a major impact on heat transfer. In one incident, a station lost over 50% of generating capacity
due to backpressure problems resulting solely from a biofilm [6]. In many cases, biofilms will
entrap fine particulates, which may further reduce heat transfer.
As mentioned previously and as shown in Fig. 4, certain types of bacteria produce corrosive
metabolic by-products (sulfate-reducing bacteria, SRBs); others (iron-oxidizing) may actually
consume base metal. Corrosion by bacteria is usually not a serious problem except in cases
where stagnant conditions exist (under deposits or scale, or off-line during unit outages). One
exception is manganese pitting of stainless steel, which occurs on-line with a thin manganese
oxide scale that is believed to be deposited by micro-organisms (Fig. 5).
Debris and/or Macrofouling
Macrofouling of the condenser inlet tubesheet and tubes can occur by any substance whose size
is close to, or greater than, tube internal diameter. Rocks, concrete pipe debris, cooling tower
materials (plastic fill, wood), chunks of ash, large pieces of rusted steel, coal, windblown debris
(paper trash, leaves or other vegetation), aquatic animals (crayfish, fish, clams), and any other
substance that enters the circulating water can obstruct cooling water flow.
Partial flow blockage of condenser tubes can occur when a tube inlet is partially covered by
debris (e.g., a piece of plastic), or when an object lodges within a tube (e.g., a rock or piece of
wood). Two damaging effects can result: (1) Slower flow through the tube will permit fine
particulates to accumulate and settle, and (2) Local flow around the lodged item may be high,
causing erosion-corrosion (primarily a problem for copper alloys). Screens are employed prior
to the condenser, with mesh size selected to prevent debris larger than tube internal diameter
from traveling to the waterbox. However, clogged screens may overflow, corrosion may create
wider openings in the screens, elongated objects may pass through screens and lodge sideways
on tubesheets, and sometimes open circulating water systems can permit entry of debris after the
screens, or debris may originate downstream of the screens (e.g. deteriorating concrete tunnels or
waterbox epoxy coatings, rust chunks).
Corrosion Fouling
Saxon [7] reported that properly cleaned copper alloy tubes in waters of low corrosivity are not
normally susceptible to pitting. Copper alloys passivate by forming a protective cuprous oxide
(Cu2O) film on the surface. This basic protective film is non-porous, very thin, and transfers
heat efficiently. Undisturbed, this film keeps the tube from pitting. Unfortunately, normal
operating conditions can be damaging to this film. Copper oxides will continue to grow in
oxygenated water of high conductivity, and a very thick outer layer of porous cupric oxide (CuO)
can develop. This porous, non-protective oxide disturbs the cuprous oxide film and concentrates
254 10937790
corrodents that can attack and pit the base metal. Thus, cupric oxide not only reduces heat
transfer, but can also accelerate corrosion. Where thick cupric oxide films develop, they must be
removed regularly to prevent pitting and extend tube life. Under normal circumstances, stainless
steels and titanium form very thin passive oxide films that neither inhibit heat transfer, nor
promote corrosion.
Prevention of Tube Interior Fouling
Suitable approaches exist for addressing routine condenser tube fouling problems. However,
applying preventive measures for all possible fouling scenarios would be cost-prohibitive,
particularly for those issues a station considers itself unlikely to encounter. Chemical and
mechanical preventive approaches are commonly utilized.
Chemical Treatment
Several chemicals, often in combination, are used to control condenser tube fouling. Chemicals
are primarily utilized with recirculating cooling towers, because (1) The concentration of
dissolved constituents is significant, increasing the threat of scaling and corrosion, and (2) Once-
through cooling systems often discharge directly into a river or lake, with chemicals restricted in
the effluent. Table I identifies some chemical treatment methods commonly used.
Treatment Method Description
pH control Most scales formed from natural waters
have greater tendency to occur at
elevated pH. Sulfuric acid is the most
common chemical used for pH
reduction, because of its comparatively
low cost and widespread availability.
scale inhibitors Phosphonates, such as HEDP (1-
hydroxyethylidene-1,1-diphosphonic
acid), and polymers such as
polyacrylate are commonly used to
inhibit scale formation, along with a
variety of related specialty molecules.
dispersants Dispersants act much like soap, keeping
fine particulates in suspension, or
causing settled particulates to go into
suspension.
biocides Oxidizing biocides are added primarily
to kill bacteria and algae, thus
preventing them from gaining a
foothold. Non-oxidizing biocides are
also frequently used.
corrosion inhibitors Zinc and phosphate inhibit corrosion of
carbon steel, and triazoles inhibit
copper-alloy corrosion.
TABLE I
Common Chemical Treatment Methods
Another common means of minimizing chemically-influenced problems is to minimize the

concentration of dissolved solids in the system, particularly in cooling towers. A balance must
10937790
255
be determined between minimizing water use, and avoiding excessive buildup of scaling and
corrosive constituents in recirculated water.
Some scales can be removed on-line (or off-line) by lowering circulating water pH to the point of
scale dissolution by increasing the sulfuric acid feed (pH of 5.8 or less for calcium carbonate, 4.0
for calcium phosphate); however, potential risks to base metal and carbon steel piping must be
carefully considered. In the case of an on-line chemical cleaning by pH reduction, one major
disadvantage is uncertainty regarding the composition of the foulant.
Mechanical Fouling Prevention
A number of approaches have been developed to prevent condenser tube fouling by mechanical
means. As previously mentioned, screens are installed to block large debris from reaching the
condenser. Settling ponds and/or clarifiers remove particulate and some dissolved constituents.
Filtration systems, such as sand filters, can also be employed to remove fine particulates, on
either a full-flow or sidestream partial flow basis. A number of systems have been developed
which send cleaning objects such as sponge balls, brushes or plastic scrapers through tubes with
cooling water flow, theoretically wiping tubes clean; some of these cleaners are retrieved at the
condenser outlet and returned automatically to the inlet, others are returned to the inlet via
backwashing. Because of the random distribution of some of the cleaning devices not all tubes
receive consistent cleaning action. Increasing cooling water flow through the condenser can be
effective in limiting microbiological growth; this will also keep temperatures lower, possibly to a
point at which scale formation is less likely. However, higher flow rates can cause increased
rates of erosion-corrosion in copper alloys.
Off-Line Options for Cleaning Fouled Tubes
Mineral scale can generally be removed by a chemical solvent (e.g. hydrochloric acid for
calcium carbonate, hydrofluoric acid for silicon-based scales). Major risks include incomplete
scale removal, in which case pitting may occur under the remaining scale where the cleaning
solution was entrapped and not adequately rinsed, and corrosion of the base metal. Additionally,
the work can be expensive, job duration may increase outage time, safety concerns exist, and the
subsequent disposal of the waste chemical solution may be costly. It has been frequently found
that some residual material must be removed by mechanical cleaning methods following the
chemical clean.
Mechanical Removal
Several viable options are used for removing condenser tube foulants mechanically. High
pressure water (8,000 – 40,000 psi) may be effective, although some potential exists for cutting
softer copper alloys. Sponge balls may be sent through tubes to remove existing foulants;
abrasive balls are available for harder foulants such as scale, although effectiveness may be
limited. Brushes of various types may be propelled through tubes; mechanically-powered
rotating brushes can also be used. Metal or plastic scrapers powered by pressurized water are
commonly used and are effective for general purpose foulant removal, plastic being used for
only the lightest deposits.
256 10937790
Large debris blocking tubes may be removed by backflushing or any of the mechanical methods
listed above; however, if debris is firmly lodged in the tube, it must be rodded out with steel bar
stock or a flexible fiberglass rod. It is generally necessary to open the waterbox and physically
remove large debris that has accumulated and blocked flow at the inlet tubesheet, or it will
simply return to the tubesheet when flow is resumed.
Off-Line Mechanical Cleaning Options
While cleaning and preventive measures for condenser tube fouling can be performed on-line,
the remainder of this paper focuses on cleaning technology innovations and procedures that are
performed off-line; specifically, mechanical cleaning options, which are the most frequently
chosen, most generally applicable and effective, and fastest cleaning methods available.
Saxon and Putman [8] addressed the benefits of mechanical cleaning for improving plant
performance and, consequently, reducing CO2 emissions. Their conclusions are straightforward:
Off-line mechanical cleaning is especially useful where fouling problems exist that are too
severe to be handled by any of the other methods. Obviously, the tool selected must be the most
appropriate for removing a particular type of deposit.
Although the use of high-pressure water can be effective with certain foulants, the jet nozzle
must be moved along the tube slowly, and the time required to clean a heat exchanger can be
excessive. Great care must be taken to avoid damaging the tubesheet, tube coatings, or inlet-end
inserts which may be present. Otherwise, the successful removal of fouling deposits may be
accomplished at the cost of new tube leaks or increased tubesheet corrosion, only recognized
after the unit has been brought back on-line. Molded plastic cleaners (pigs) are quite popular for
some light silt or microbiological film applications. Brushes can also be used to remove these
soft deposits. Brushes are also useful for cleaning tubes with enhanced surfaces (e.g. spirally
indented or finned), or those tubes with thin wall metal inserts or epoxy type inlet coatings.
Metal Scrapers
With harder types of deposits, calcium carbonate (calcite) being a notable example, metal
cleaners of various designs have been developed for effective removal. The spring-loaded blades
of the metal cleaner are an essential element in the success of the technology. The use of spring-
loaded tube cleaners was identified as a best maintenance practice by Putman and Walker [9].
The blades are mounted on a spindle (Fig. 6). At one end of the spindle has a serrated plastic
disk that allows a jet of water to propel the cleaners through a tube with greater hydraulic
efficiency. The water is directed to the tube being cleaned by a water gun, and is delivered by a
portable pump. Water pressure of 300 psig is most effective for propelling the cleaning tools
through the tubes at their design travel speed of ten to twenty feet per second. The use of water
pressure also prevents the cleaner’s exit velocity from rising above a safe level. Some other
cleaning systems use air or a mixture of air and water to propel the cleaner, but air pressure is
compressible and dangerous to use. Since the pump is usually mounted on a wheeled base plate,
the system can be conveniently moved from unit to unit within a plant, or even moved to another
plant.
10937790
257
Another advantage of using water for tube cleaner propulsion is that the material removed is
rinsed out, and can be collected in a plastic container for later drying and weighing to establish
2
the deposit density (g/m ), followed in many cases by laboratory analysis of the deposit.
Most metal cleaners are designed to have a controlled spring-loaded cutting edge; however, if
effective deposit removal is to be the result, the dimensions of the cutting surfaces have to be
closely matched to the internal diameter of the tube being cleaned. This not only improves the
peripheral surface contact, but also ensures that the appropriate spring tension will be applied as
the cleaner moves through the tube (Fig. 6). The effective life of cleaners with this design can be
as high as 12 tube passes.
In practical experience, Kim [10] found that annual off-line mechanical tube cleanings using
metal scrapers was effective in reducing the corrosion rate, retarding pitting and increasing unit
availability, reversing a trend of increasing tube leaks and plugging.
Innovations in Cleaning Technology
As a result of an innovative research program organized to resolve problems encountered in the

field and to develop new products where existing equipment was found to be inadequate, new
tube cleaners have been developed. For example, in order to provide the blades with more
circumferential coverage of the tube surface, a new scraper design was introd (Fig. 7). The
increased contact surface provided by the greater number of blades was found to be more
efficient in removing tenacious deposits such as those consisting of various forms of manganese.
Another development involved a tool for removing hard calcite deposits, which were sometimes
found to be difficult to remove in a timely manner, even by acid cleaning. This cleaner is shown
in Fig. 8, and consists of a teflon body on which are mounted a number of rotary cutters. These
are placed at different angles around the body, which is fitted with a plastic disk similar to those
used to propel other cleaners through tubes. Used on condenser tubes that had accumulated a
large quantity of very hard deposits, Stiemsma, et al [11] described how cleaners of this type
removed 80 tons of calcite material from a large surface condenser. It has now become a
standard tool whenever hard and brittle deposits are encountered. Hansen and Saxon [12]
demonstrated the practical use of this tool for removing scale from stainless steel condenser
tubes. Boroscopic photography shows the effectiveness of this cleaning tool in Fig. 9 and Fig.
10.
Additional developments for the removal of manganese dioxide, iron oxide, and silica deposits
include the stainless steel brush, which has over 1,000 contact points per brush (Fig. 11). Actual
aluminum brass tube samples with iron and manganese deposits were examined in the “as found”
and “as cleaned” condition (Fig. 12). After one pass of the stainless steel brush the cleanliness
factor for the tube from waterbox A went from 83.3 percent to 91.5 percent, and from 73.6% to
94.1% in the tube from waterbox B.
The experience gained from using these techniques has allowed the cleaning duration to be
forecast with confidence and cleaning to be performed on schedule. For instance, a normal crew
of four technicians can clean 5,000 tubes during a 12-hour shift. Clearly, the number of tubes
cleaned in a day could rise with an increase in crew size, limited only by the availability of
adequate space for the crew to work effectively.
258 10937790
All off-line cleaning methods may sometimes require additional deposit removal assistance,
especially where the deposits have been allowed to build up excessively, and sometimes become
harder. In such cases, for example, it may be necessary to acid clean, followed by cleaning with
mechanical cleaners or high-pressure water to remove any remaining debris. Tube cleanliness
can then be maintained by regularly scheduled off-line cleaning.
Effect of Metal Scrapers on Tube Surface
One perceived drawback of using metal scrapers to clean condenser tubing is their effect on tube
metal, including concerns over incompatibility of the carbon steel scrapers with various tube
metal alloys, and the possibility that base metal might be removed by the scraper digging into the
tube wall. Additionally, many plant operators are concerned that metal scrapers will remove
protective oxide coatings from tube surfaces, thus exposing base metal to accelerated local
corrosion.
With cleaners that have been properly designed and carefully manufactured, loss of base metal is
very minor. Indeed, Hovland, et al [13] tested such carbon steel cleaners, manufactured by
Conco Systems, Inc., by repeatedly passing them through 30 feet long 90-10 CuNi tubes. It was
found that after 100 passes of these cleaners, wall thickness reduction was only between 0.0005
and 0.0009 inches (12.5 and 23 µm). At this rate, extrapolating the worst-case removal from this
series of tests for a 0.049” wall thickness, it would take about 2700 passes of a cleaner per tube,
or roughly 1000 years of annual condenser cleanings to result in a critical reduction in tube wall
loss.
Regarding exposure of base metal to accelerated corrosion due to removal of a protective oxide
layer, no evidence of this scenario has been produced. In fact, where protective oxide is
removed, it is rapidly re-formed upon exposure to oxygenated water, with no detrimental effects.
Galvanic Corrosion of Stainless Steel
According to electrochemical theory, metals and alloys can be ordered according to relative
reactivity, with magnesium, zinc and aluminum towards the active end of the series, and
titanium, passive stainless steel and copper alloys towards the passive or noble end. When a
galvanic couple exists (two different metals or alloys in physical contact), the less noble partner
acts as an anode and will corrode (oxidize). The relative surface areas of anode and cathode are
also important in determining the corrosion rate; if a large cathodic area is associated with a
small anodic area, the corrosion rate is increased, since all of the electrochemical oxidation
current is focused on a small region. In general, the corrosion rate is also increased with
increasing water salinity (dissolved solids) and / or increasing temperature.
Note that stainless steel can be in either an active or passive state. In air, stainless steel is
passive. Immersed in water, passive stainless steel is more noble than copper alloys, and much
more noble than carbon steel. Its normal state in oxygenated fresh water is passive, provided the
tube surfaces are clean. In saline water, anodic protection may be required to maintain the more
noble (passive) state, in order to reduce the rate of corrosion. Note that a variety of stainless
steel alloys exist, including the austenitic (300-series) and ferritic types (400-series). Specialized
stainless steels can be very corrosion resistant, including the austenitic alloy AL6XN (high Cr,
10937790
259
Ni, Mo, and N), and the ferritic alloy SeaCure (high Cr and Mo). Stainless steel alloys are
susceptible to pitting by under-deposit corrosion, as they depend on a thin surface chromium
oxide layer for passivation, and this layer is best maintained by continuous exposure to oxygen.
Most stainless steel condenser tube installations have been composed of the 304 or 316 alloys.
They were first installed for surface condenser applications over 45 years ago, and have seen
many routine cleanings with both carbon steel and stainless steel scrapers. As described by
Saxon [14] and later Putman [15], it is clear that these tubes, as with other alloys, frequently
require cleaning.
While concerns over material incompatibility have been expressed, in practice there has been no
evidence of carbon steel particles from properly designed mechanical scrapers embedding into
stainless steel tubes. Tests were conducted by Anderson [16] on stainless steel tubes that had
been cleaned with carbon steel mechanical scrapers. No transfer of carbon steel onto the
stainless steel tube surfaces was observed, and this has been the common experience. There is
no possibility of galvanic corrosion of the stainless steel; embedded carbon steel particles would
corrode preferentially. After such particulates are oxidized to iron oxide they should be
removed, similar to any other particulate fouling. It is well known from the literature and in
practice that if 300-series stainless steel tubes are not kept clean, deposits such as naturally
occurring iron oxide, manganese oxide, calcium carbonate and bacterial accumulations are prone
to cause pitting and crevice corrosion. Stainless steel condenser tubes must be kept clean, and
properly-designed metal scrapers, either carbon steel or stainless steel, are an effective option.
While this section has focused on 300-series stainless steel tubes, similar considerations would
apply to the possible embedding of carbon steel scraper material in other common tube alloys,
such as titanium and copper alloys – the carbon steel debris would oxidize preferentially to the
base metal.
Selecting a Cleaning Procedure
Regardless of the tube material, the most effective way to ensure that tubes achieve their full life
expectancy and maximum heat transfer efficiency is to keep them clean. Each time the tube
deposits, sedimentation, biofouling and obstructions are properly removed, the cleaned surfaces
should be returned as nearly to bare metal as possible. This restores optimal heat transfer and
essentially supplies the tube itself with a new life cycle, as protective oxide coatings quickly
rebuild themselves to re-passivate the cleaned tube.
While most existing foulants can be removed by either chemical or mechanical means, selection
of an optimal method is generally based on the type of foulant and the cost of the technique.
Putman [17] described the important criteria in selecting an appropriate cleaning procedure: the
selected cleaning procedure should remove the particular deposits that are present as effectively
as possible, and render the unit out of service for the minimum amount of time. Some other major
considerations in the selection process are discussed briefly in the following paragraphs.
Removal of Obstructions
Many tube-cleaning methods are ineffective when there are obstructions within tubes, or when
various forms of macrofouling are present. When such obstructions are found it is inadvisable to
proceed with the cleaning regimen as planned. Attention should be given to shellfish which
260 10937790
constitute macrofouling, and can include Asiatic clams and zebra mussels in cooling water.
Other obstructions generated directly from the process medium must be considered as well. The
selected tube cleaner must have the body and strength to remove such obstructions. The cleaning
method may be required to remove the byssal material that shellfish use to attach themselves to
tube walls, as well as hard, aged deposits.
Certain types of other debris and process impurities can become obstructions, including cooling
tower fill, waste construction material, sponge rubber balls, rocks, sticks, seaweed and fresh
water trash, any or all of which can become lodged in the tubes or on the tubesheet, and will
require removal. Cleaning objects can themselves become obstructions, although experience has
shown that if appropriate procedures are followed, properly designed cleaners will not become
stuck inside tubes, unless the tube is deformed.
Removal of Corrosion Products
When heat exchangers are equipped with copper alloy tubing, corrosion product growth under
deposits is a particular concern. Copper oxide corrosion products can grow to the point at which
they will seriously impede heat transfer. Not only will the performance of the condenser be
degraded, but such deposits will also increase the potential for tube failure. Under deposits, a
protective inner cuprous oxide film cannot be maintained; thus, the base metal is exposed to
attack. When destructive copper oxide accumulations occur, they must be removed, together
with overlying deposits and scale.
Surface Roughness
Rough tube surfaces, often the result of accumulated fouling deposits, are associated with
increased friction coefficients, while the reduced cooling water flow rates allow deposits to
accumulate faster. It has been found that rough tube surfaces pit more readily than smooth
surfaces, and are also responsible for inducing erosion-corrosion damage in copper-alloy tubes
[18]. A tube surface rendered smooth from effective cleaning can improve condenser
performance in several respects:
• Lower water temperature rise across the heat exchanger, reducing the heat loss to the
environment (improved heat transfer capacity)
• Increase in both flow volume and water velocity, often resulting in reduced pumping
power requirements
• Increased time required between cleanings, by reducing rate of re-deposition of fouling
material on the tube surfaces.
• Reduced pitting from turbulence and gas bubble implosion
• Longer tube life
Conclusion
Amidst a marketplace replete with numerous cleaning and service options for removing fouling
from the interior of condenser tubes, site engineers and staff must choose maintenance practices
which will help to improve performance while protecting from corrosion, ensuring that the
integrity of their equipment will not be compromised. Innovative techniques or procedures may
10937790
261
be required. Interim gains in performance can be achieved while maintaining the long-term
efficacy of a given unit, so long as the cleaning technology applied is sound and site-specific.
References
th
1. TEMA (1988). Standards of the Tubular Heat Exchanger Manufacturers Association,7 ed.,
Tarryton, NY.
rd
2. HEI (1998). Standards for Power Plant Heat Exchangers, 3 Edition, Heat Exchange
Institute, Cleveland, OH
3. Putman, R.E., Steam Surface Condensers: Basic Principles, Performance Monitoring, and
Maintenance, 2001 ASME Press, New York, NY, pp 172-174, 183,184.
4. Saxon, G.E.,Jr., R.E. Putman, and R.Schwarz (1996), “Diagnostic Technique for the
Assessment of Tube Fouling Characteristics and Innovation of Cleaning Technology,”
Proceedings, EPRI Condenser Technology Conference, Boston, August, pp 14-1 to 14-14.
5. Howell, A.G., and Schwarz, R.C, “Erosion-Corrosion in Copper-Alloy Condenser Tubes,

Proceedings, EPRI Condenser Technology Conference, St. Petersburg, FL, September 1993.
6. Howell, A.G., Internal Report, 2004.
7. Saxon, G.E., “Copper Alloy Tube Cleaning,” TPT-, Vol. 1, No. 1, September – October,
1990, Publ. INTRAS Inc., Danbury, CT., pp. 28-30.
8. Saxon, G.E., Jr., and R.E. Putman (2003), “Improved Plant Performance and Reduced CO2
Emission Through State-of-the-Art Condenser Cleaning and Air-Inleakage Detection”
Proceedings, ASME IJPGC 2003, Atlanta, GA, June 16-19, cd
9. Putman, R.E. and Walker, R., “Proper Maintenance Practices Involving Condenser Cleaning
and In-leakage Inspection”, Proceedings of the International Conference on Power Station
Maintenance 2000, Oxford, England, The Institution of Mechanical Engineers (IMechE),
Sept. 18-20, 2000, pp. 161-170.
10. Kim, D.S. and Putman, R.E. (1993),“The Management of Condenser Tube Pitting at #2 Yong
Nam Thermal Power Plant to Improve Unit Performance and Availability,” Proceedings of
IJPGC, Kansas City, October 1993, PWR-Vol.23, pp.43-48.
11. Steimsma, R.L., Bhayana, G.K. and Thurston, R.D., (1994),. “Performance Enhancement by
an Innovative Tube Cleaning Application,” Proceedings of IJPGC 1994, Phoenix, AZ,
American Society of Mechanical Engineers, (ASME) New York, PWR-25-1994, pp. 7-11.
12. Hansen, J.T and Saxon,G.E.,Jr., “Improving Condenser Efficiency with Innovative Scale
Removal System Technology”, Proceedings of Power 2004: ASME Power 2004, Baltimore,
Maryland, March 30 – April 4, 2004
262 10937790
13. Hovland, Alan W., Rankin, Drew A., and Saxon, Edward G., “Heat Exchanger Tube Wear
by Mechanical Cleaners,” Proceedings of The Joint Power Generation Conference, 1988,
Philadelphia, PA, Sept. 25-29, 1988.
14. Saxon, G.E., Stainless Steel Tube Cleaning, Conco Bullets, Vol. 3, Issue No. 7, 1988. pp 1-2.
15. Ibid, Putman, R.E., 2001.
16. Anderson, Susan et. al., “Mechanical Manganese Dioxide Removal in Stainless Steel
Condenser Tubes,” EPRI Service Water Systems Cleaning Technology Seminar, Palo Alto,
CA, February 21-22, 1990.
17. Ibid, Putman, R.E., 2001.
18. Ibid, Howell, A.G. and Schwarz, R.C., 1993.
COOLING WATER
+ - + -
H OH Cl - H OH
Cl -
Cl - Cl - Cl - Cl - O2
O2 O2 O2
POROUS
DEPOSIT
METAL
OXIDE
e- e-
+ Cl - TUBE MATERIAL
H
SHELL SIDE
Figure 1
Model of Under-deposit Corrosion
10937790
263
Figure 2
Pitting Under Scale in Stainless Steel Tube
Figure 3
Erosion-corrosion Around Deposits
Figure 4
Bacteria Consuming Base Tube Metal
264 10937790
Figure 5
Pitting Due to Manganese Deposits
Tube Wall Cleaning Blade Deposits
Figure 6
Tube Cleaner in Action
(Figure Courtesy of Conco Systems, Inc.)
Figure 7
Hex Cleaner
Patent No. 0698423
(Photo Courtesy of Conco Systems, Inc.)
10937790
265
Figure 8
Cal Buster
Patent No. 5153963
Figure 9
Stainless Steel Condenser Tube with Calcium Carbonate Scale
Figure 10
Stainless Steel Condenser Tube with Calcium Carbonate Scale Removed
266 10937790
Figure 11
Stainless Steel Tube Cleaning Brush
Patent No. D450035
A Water Box As Found Tube

CF 83.3%
Cleaned Tube
SSTB 1 Pass
CF 91.5%
B Water Box As Found Tube

CF 73.6%
Cleaned Tube
SSTB 1 Pass
CF 94.1%
Figure 12
Aluminum Brass Tubes Tested for Manganese and Iron Deposit Removal
10937790
267
10937790
Technology for Cleaning the
External Surfaces of
Air-Cooled Condensers
Gary Fischer
Conco Systems, Inc.

August 30 – September 1, 2005
San Diego, CA
10937790
269
Air Cooled Condensers
Air cooling may be the only practical condensing technology
for certain sites
Mine-mouth power plant with inadequate local cooling

water source (Black Hills, WY)
Plant situated in a desert (El Dorado, NV)
Cooling tower plume and fog would endanger highway
safety (Wyodak, WY)
Thermal pollution with once-through system must be
avoided (Athens, NY)
Conventional cooling towers intrude on rural landscape
or degrade a residential area, making a permit difficult to
obtain
270 10937790
Installed cost tends to be more

expensive than their water-cooled
equivalents
An energy penalty can be incurred
during summer conditions
EPA does not consider air cooling to be
the Best Available Technology (BAT)
10937790
271
Features of
A-Frame Construction
Finned tubes
Parallel Flow Condensing Sections
Counter Flow Condensing Sections
Fans located in Base of A-frame
Lengthy and large diameter exhaust piping
prone to air inleakage
272 10937790
Air Cooled Condenser
10937790
273
Parallel Flow Panels
TURBOGENERATOR
EXHAUST VAPOR
INLET
HEADER
PARALLEL PARALLEL
FLOW FLOW
CONDENSER CONDENSER
VAPOR
AIR AIR
FLOW FLOW
CONDENSATE
MOTOR
FAN
VAPOR FLOW
CONDENSATE COLLECTOR
CONDENSATE RETURN
AIR-COOLED CONDENSER
PARALLEL FLOW SECTIONS
FIGURE 2.0
274 10937790
Deaerator Panels
TURBOGENERATOR
EXHAUST VAPOR
STEAM JET INLET

STEAM JET
AIR EJECTOR HEADER
AIR EJECTOR
CONDEN- VAPOR
SATE
AIR COUNTER AIR

FLOW FLOW FLOW
CONDENSER
(DEPHLEGMATOR)
(OR DEAERATOR)
MOTOR
FAN
VAPOR FLOW
CONDENSATE COLLECTOR
CONDENSATE RETURN
DEAERATOR SECTION
FIGURE 3.0
10937790
275
Fouling Tendencies
Finned tubes tend to collect pollen, dust,
insects, plastic bags, bird carcasses, etc.
Reduced air flow rate reduces heat transfer
which increases heat rate and/or reduces
generation capacity
Water, sometimes sprayed on tubes during high
ambient temperature periods, can lead to the
formation of scale on tube fins thus reducing the
heat transfer rate
276 10937790
Fouled External Surface
10937790
277
Impact of Fouling
Poor heat transfer

Higher operating costs
Increased power supply
of fandrive motors
Deterioration of turbine
back pressure
Restricted MW output
278 10937790
Operating Criteria
Ambient Air Temperature
Air temperature rise across tube
banks
Pressure drop across tube
banks
Pressure rise across fans
Condenser back pressure
Condenser Duty:
Q = Wcond (Hvap – Hliq)
10937790
279
Instrumentation
Ts
EXHAUST VAPOR
Condenser back pressure

INLET
Air temperature rise across tube banks HEADER

Condenser Duty:
T ao
Q = Wcond (Hvap – Hliq)
Pressure drop across tube banks

T ai
p
ao
p
ai
MOTOR
P fo
FAN
Ambient Air Temperature P

fi
Tamb Pressure rise across fans
INSTRUMENTTAION
FIGURE 4.0
280 10937790
Operating Data Correlation
ASME is moving forward with performance
standards and an industry draft is expected
by year end.
Two common ways of presenting air cooled
condenser performance data:
– Expected condenser duty plotted against inlet dry
bulb air temperature for various values of
condenser back pressure
– Expected condenser back pressure plotted
against percent design air flow for various values
of ambient air temperature
10937790
281
Cleaning Techniques
Three principal methods of cleaning an
air cooled condenser:
Fire hose
High pressure handlance
Automated cleaning machine
282 10937790
Fire Hose
High volume of water but low washing
effect
Unit must be taken out of service and
scaffolding erected
Improvements are quite small, since only a
portion of debris is removed, remainder
being compacted between tube fins
10937790
283
High Pressure Handlance
Low water consumption and a high water
pressure
Latter can damage galvanized surfaces
and/or snap off fins
Unit must be taken out of service and
scaffolding erected
Again, improvements are quite small since
only portion of debris is removed, remainder
being compacted between tube fins
284 10937790
Automated Cleaning Machine
Nozzle beam optimally matched to tube bundle
geometry with a constant jet angle
60 GPM at 1,000-2,000 psi water pressure avoids
fin and tube surface damage
Nozzle design, distance from surface and jet
energy adjustable
Carriage moves at a constant speed
Water contains no additives
Fouling removed effectively and uniformly
No need to shut unit down or erect scaffolding
10937790
285
Permanently Installed System
286 10937790
Nozzle Beam
10937790
287
Nozzle Satisfies Fin Geometry
288 10937790
Semi-Automated System
10937790
289
Portable System
290 10937790
Cleaning Results
Fouled
During Cleaning
Cleaned
10937790
291
Performance Improvements
Cleaning almost invariably allows fan speeds to
be reduced, reducing auxiliary power
consumption
In some plants, cleaning can also result in
increase in generation capacity (e.g. from 15
MW to 18 MW)
Economic savings from cleaning can be
estimated using simple calculations. In one UK
plant it was estimated to be $18,476 /week
292 10937790
Economic Efficiency Calculation
for a 400 MW power plant
8 mbar ≈ 1 MW ≈ $45.00
Before cleaning turbine back pressure 130 mbar

After cleaning turbine back pressure 100 mbar
Difference 30 mbar ≈ 3.8 MW
Increase in weekly revenue after cleaning:

3.8MW x 45.00 $ x 7 days x 24h x 75% running time =
$21,546.00 $/week
10937790
293
Conclusions
Air cooled condensers are a viable alternative to
steam surface condensers.
They allow a plant to be built on sites that are
otherwise subject to impossible design
constraints.
Because of the fouling tendencies air cooled
condensers require effective cleaning systems.
Performance improvements may be achieved by
maintaining the cleanliness of the external
surfaces.
294 10937790
5
SESSION 5: PLENARY SESSION II
10937790
295
10937790
ALSTOM Power
April, 2005
Vibration Prevention Goes

Beyond Support Plate Spacing
David H. Cooley – V.P. Technology

Edward K. Yost – Supervising Application Engineer
ABSTRACT
In the early 2000’s Alstom received orders for two(2) 4 x 550 MW combined cycle
power plants. One of the power plants was located in the state of Arizona and
the other in Arkansas. Each generating unit consisted of two (2) gas turbines
and HRSG’s supplying steam to one tandem compound two flow steam turbine.
Each condenser was two pass vertically divided with 165,000 sq. ft. of
condensing surface. Because of the water quality one jobsite had 1” OD – 22
BWG titanium tubes and the other jobsite had 1” OD – 22 BWG type 316
stainless steel tubes.
Within the first few hours of operation the first unit with titanium tubing had
fatigue failures occur. This paper examines the cause of the failure and provides
an insight to the difference in the vibration resistance of titanium and stainless
steel.
CONDENSER DESIGN PARAMETERS
The condenser design had to consider three different operating cases. Case 1 is
normal operation, Case 2 is normal start up and Case 3 is full load rejection.
Operation Cases
Case 1 Case 2 Case 3
Steam Turbine Flow (lb/hr): 1,399,923 556,732 N/A
IP Dump (lb/hr): 1,029,000 2,058,000
LP Dump (lb/hr): 70,000 140,000
% Increase in Tot. Steam Flow: 127% 171%
10937790
297
The condenser also had 238,500 lb/hr of steam dome hood desuperheating
spray flow that added to the total steam flow in calculating the percentage
increases.
Case 3 flow is an event that will occur only if the steam turbine trips at full load.
The IP dump flow is comprised of 1,500,000 lb/hr from the gas turbines at full
load, 468,325 lb/hr from both duct burners in operation and the remainder is a
4% safety margin for a total of 2,058,000 lb/hr. When the steam turbine trips the
duct burners are tripped and the flow decays quickly to the 1,500,000 lb/hr figure.
The, as built, condenser support plate spacing is 29 ½”. The Heat Exchange
Institute (HEI) Condenser Standards in effect at the time of design would allow a
maximum mid span spacing of 33” for titanium and 38” for the stainless steels.
Alstom’s internal design requirements require a spacing that limits the minimum
tube natural frequency to 70 Hz. At the existing spacing the tube natural
frequency is 72 Hz for titanium and 86 Hz for stainless steel. The HEI issued an
addendum in Dec. 2002 and the as built support plate spacing also meets these
requirements. It is apparent that the support plate spacing correctly meets both
these standards.
Alstom also evaluates the tube bundles for stiffness against fluid elastic whirling
using the Connors method. Connors defines the critical velocity (vc) as the
velocity at which fluid elastic whirling is initiated. The Connors method is applied
by comparing the density (σ) times the velocity (v) squared (σv2) at the design
operating conditions to the critical conditions (σv2c). The ratio of σv2 / σv2c is
called the vibration risk factor (R). It can be seen that “R” must be less than one
to prevent vibration.
Prudent design requires some safety factor for unknowns and uneven steam
distribution. Alstom uses the following guidelines for design. For normal
operation the design “R” should not exceed .5 and for transient operating
conditions the “R” should not exceed .7. These “R” values are calculated on
average velocities at the severest mode of operation. The severest mode is
always coldest water and highest load with 100% clean tubes.
The “R” factor must be evaluated at the highest velocity point in the condenser.
In this design the highest velocity point is at the top entrance to the ducts that
distribute the steam between the tube bundles and between the tube bundles
and the condenser sidewalls.
298 10937790
In evaluating steam dump cases (i.e. 2 & 3) the specific volume increase from
the superheat was accounted for. Also the flow increase from the evaporated
dome hood sprays was added to the total flow.
The following table provides the “R” factors based on average velocities for the
lowest water temperature of 55°F, the guarantee temperature of 86°F and an
intermediate temperature of 70°F.
Stainless Steel Tubes Operation Cases

Press. In. HgA @ 86°F: 2.43 2.94 4.03
"R" factor: 0.23 0.29 0.34
Press. In. HgA @ 70°F: 1.55 1.91 2.70
"R" factor: 0.28 0.37 0.42
Lowest Press. In. HgA @ 55°F: 1.02 1.30 1.92
Severest "R" factor: 0.35 0.45 0.51
Titanium Tubes Operation Cases

Press. In. HgA @ 86°F: 2.38 2.87 3.90
"R" factor: 0.31 0.40 0.47
Press. In. HgA @ 70°F: 1.51 1.86 2.60
"R" factor: 0.39 0.50 0.58
Lowest Press. In. HgA @ 55°F: 0.99 1.25 1.83
Severest "R" factor: 0.48 0.62 0.70
You will note that at the severest operating condition the vibration risk factors for
the normal operating Case 1 meet the .5 maximum and for the other operating
condition Case 3 meet the .7 maximum. The difference in the risk factors for the
different tube materials is due entirely to the difference in the modulus of
elasticity of the materials. The titanium modulus is 55% of the stainless steel
modulus.
It can thus be concluded that the condenser is properly designed, in accordance

with Alstom and HEI standards, and that the support plate spacing design is not
the root cause of the failures.
10937790
299
FAILURE ANALYSIS
The first of the titanium units was in initial start up. It was operating in steam
dump mode and within approximately 14 hours tube failures occurred. At the
time of the titanium tube failures the stainless steel tube condensers had been in
operation for about 6 months with no failures.
The failures occurred on the left hand side of the left hand tube bundle in the duct
between the tube bundle and the condenser side wall. The following illustration
indicates the tubes that failed.
Failed
Tubes
The plant DCS system indicated at the time of tube failure the condenser was
seeing 1,250,000 lb/hr of total dump flow and 77.4°F cooling water. These are
the conditions used for the failure analysis. The flow was proportioned between
the IP and LP steam dumps in accordance with their original design flows for
thermal calculations.
Minimization of the vibration risk factor in a condenser is directly proportional to

the velocity squared for any condensing condition. The original design of the
condenser contemplated a steam dump piping arrangement that provided
proportional steam distribution to minimize the velocities. This arrangement is
indicated in the following sketch
300 10937790
1X 1X 1X 1X
1X 2X 1X
The large diameter lines indicate the IP bypass and the small diameter lines
indicate the LP bypass. The solid color indicates that the steam dump piping is
entering the steam dome from the inlet/outlet end and the hash line color
indicates entering the steam dome from the return end. For the remainder of the
discussion we will assume all the flow comes from the IP bypass since the LP
represents only 6% of the total flow.
The short fat yellow arrows indicate the design capacity down the ducts at each
side of the tube bundle. Note the center duct normally handles 2X the flow since
it is twice the width and must feed the side and the bottom of the two adjacent
tube bundles.
The long skinny red arrows indicate the steam discharge direction of the steam
dump. The length of the steam discharge arrows is in proportional to and
indicative of the EPRI minimum safe distance. Each 1X indicates 25% of the
steam dump flow. The arrangement of the steam dump pipes provides even
distribution across the width of the condenser in proportion with the duct
capacities to minimize the velocities.
10937790
301
As the design phase of the condenser contract progressed the steam dump line
location was changed to accommodate the plant layout. The final end view of the
steam dump arrangement is depicted in the following sketch.
1X 1X
2X
1X 1X
1X 2X 1X
Des. Des. Des.
Area of
Failed
Tubes
The arrangement “as constructed” provides a steam dump distribution that is not
in proportion to the duct design capacity. The left side of the condenser receives
about twice the amount of steam dump as the duct has design capacity for.
Viewing the condenser from the longitudinal direction we know that because of
the slope of the end panels of the steam dome that the upper steam dump pipe is
shorter than the lower one. This difference means that the distribution of the
steam dump will be concentrated more in the center of the condenser. The
estimated steam distribution longitudinally was based on the “as constructed”
steam dump pipe arrangement and is shown in the following graph.
302 10937790
Estimated Steam Dump Distribution to Norm
350%
300%
250%
200%
150%
100%
3 4 5 6 7 8 9 10
Condenser Bay Number from Inlet/Outlet End
The net result is that bays 6 –9 experience higher steam flows by approximately
300%. This is sufficiently high to increase the vibration risk factor (R) to
unacceptable levels. The titanium tube “R” is greater than 1.0 indicating fluid
elastic whirling. The stainless steel tube “R” value exceeds 0.7 but remains
below 1.0.
We know from the field data that tubes were lost in the duct on the left hand side
of the left hand tube bundle in bays 4, 6, 7, 8 & 9. The following table provides
the local vibration risk factors based on the condensing pressures at 77.4°F and
1,250,000 lb/hr of steam dump flow.
BAY #: 3 4 5 6 7 8 9 10
TITANIUM "R": 0.47 0.47 0.60 >1 >1 >1 >1 0.60
S/S "R": 0.34 0.44 0.44 0.93 0.80 0.85 0.85 0.44
Tube Failures: 1 1 1 3 1
“R” factors exceeding 1.0 correlate well with the actual failures considering the
short time of operation and the small number of failures to date. The results also
provide an insight as to why no failures have been experienced on the stainless
steel units even though the operating time was considerably longer. The lone
failure in bay #4 where the risk factors are low may be the result of some steam
dump system upset or just part of the statistical distribution.
This same steam distribution will exist under all steam dump conditions including
the worse case full load trip. Thus at the full load trip the stainless steel units will
also experience vibration risk factors >1.0 and potential tube loss.
10937790
303
CONCLUSIONS
This case study is a good example, but not the only example in the industry, of
how tube vibration is a direct function of the local velocity. The base support
plate spacing provided a condenser that met industry and the manufacturer’s
standards for proper design. It would not be expected that tubes would be lost
by fatigue failure. Tube fatigue failure from improper support plate spacing
manifests itself by loss of tubes through out the entire condenser (i.e. not
localized losses).
The “as built” steam dump system layout created an uneven steam distribution
that overloaded the left side of the condenser and produced unacceptable
vibration risk factors for the titanium condenser under normal start up. The
stainless steel condenser would also have unacceptable vibration risk factors at
full load trip conditions. Staking the left side of the left hand tube bundle on all
units mitigated the problem.
The tubes were lost because of elevated local velocities driven by the steam
dump line locations that could not provide for even distribution. It is important for
both the condenser designer and the Architect Engineer to understand the
criticality of even flows on large steam dumps and to work together to provide
connection locations that promote this within the condenser.
The problem of local velocities driving vibration risk factors to unacceptable

levels is ever present with most service connections albeit on a smaller scale
than with steam dump. The condenser designer must be aware of the potential
and provide distribution devices that bring local velocities to acceptable levels.
One plant had type 316 Stainless steel tubing with a modulus of elasticity of
28,300,000 in/in2. The other plant had titanium tubing with modulus of elasticity
of 15,500,000 in/in2. It is this characteristic that makes titanium more susceptible
to fatigue failure for a given situation.
In this particular example the titanium unit failed within hours of start up and the
stainless steel unit had not experienced any failures. Thus more care must be
exercised concerning local velocities when designing condensers with titanium
tubing.
304 10937790
REFERENCES:
• Heat Exchange Institute, Inc., Standards for Steam Surface Condensers,

Ninth Edition – 1995 and Addendum 1 – 2002
• Heat Transfer Research, Inc., Report STV-1 Tube Vibrations in Shell –

and- Tube Heat exchangers - 1974
10937790
305
10937790
Development of a Cobalt-Phthalocyanine Catalytic Thin Film To Inhibit Tubeside Biofilm Adherence in
Glenn J. Tiffin, PE AmerenUE Performance Engineering (mc 613- Lab Services)

1530 Singleton St.
St. Louis, Missouri 63166 -6149
mailto:gtiffin@ameren.com
( This is an R&D steering supplement to the .ppt : “AmerenUE Copper-Beryllium Tube Trial on Labadie
Plant Unit 2”, to be presented at the EPRI* Condenser Technology Symposium, September 1, 2005 in
San Diego – adjunct to slide 14 )
Introduction
Improving the thermal performance of Electric utility industry steam surface condensers is probably the
single largest contributory influence1 to reducing overall turbine-cycle heat rate (i.e., reducing “fuel
burn”).
Towards the end of the 1990’s, many aging generating facilities created a demand2 for higher-performing
replacement tubing as original equipment condenser tubing --typically 0.049”wall Admiralty brass being
quite commonplace as the OEM’s tube-- began to either thin-through from the tubeside due to grit-
scouring, and/or else condensate-groove from the shellside as failure mechanisms in developing through-
wall leaks. Once the leaks developed, unit load reduction forced by silica excursions provided the
impetus to re-tube.
The prevailing replacement design was to phase-out use of cuprous alloys in the entire lower-heat cycle,
including condenser, to avoid, for example, also the possibility of shell-side wastage of Admiralty brass
resulting in copper deposits on turbine blades (i.e., ‘copper-carryover’, and associated efficiency losses it
incurred).
Consequently, “in-kind“ tube material selection for re-tubings was no longer in vogue—opportunities to
eradicate copper were viewed as opportunities to avoid the need to add condensate polishers. Stainless
steel became the new material of choice for re-tubings– already there originally in peripheral rows for
vibration resistance, it now quickly became the ‘copper-free’ choice for the entire tube bundle due to its’
hardness and wearability.
Problems
Other than stress-corrosion cracking in the presence of chlorides (i.e., seacoast sources of cooling water),
failure experience record with through-wall pitting (see slide 11 of the .ppt) and under-deposit corrosion
with various grades of stainless in river-cooling water use was not widely established –nor understood-
due to again, lack of heretofore widespread stainless predominance in this application (again, beyond the
peripheral row tubecount), and the varying degree of organic exposure ( indigenous river conditions )
varying greatly from site –to-site.
To make matters worse, even when tubeside biofouling is not present, without addition of heat transfer
surface (as is the case with a re-tubing vs. a modular condenser retrofit) even when the highest thermal
conductivity grade of stainless (439 grade ferritic) runs at design cleanliness factor (85%), the overall heat
transfer coefficient is still only 92% that of the original Admiralty tubing3, which can place the generating
10937790
307
unit thermally marginal during the summer months ( back-pressure limited to > 5” HGa ), especially
during low river levels when circ. water pump flow output may also be deficient.
439 stainless (was- in the late ‘90’s) a very economical replacement tube selection, but, in actual
(Missouri) River use, was found to live up to it’s ‘PREN’ value4 of only 17 in pitting resistance, and
organic under-deposit attack became evident, if not (at least) the occasional calcite scale, as well.
Figure 1
Crossroads
Additional capital investment to keep the 439 stainless tubes continually clean (a ‘sponge-ball’ system)
was placed in pilot trial-use for evaluation on one generating unit ( unit 4), but not without some setbacks
in debris-filter reliability. In addition to more escalated backflush regimens, a cleaning contractor
program was also adopted as extraordinary means to keep the tubes ‘tubercle-free’ from MIC attack. We
also had adopted ‘critter-cooks’ – i.e., steamside exposure without tubeside cooling flow , as well as more
concerted efforts to dry out the tubes during extended shutdowns , using portable blowers. Consideration
was also given to circulating water treatment (biocide injection), based on DNA analysis of bioslime
5
specimens .
All these efforts did not seem to address the fundamental issue of why the biofilm was establishing
adherence sites on the tubes in the first place, and not simply being carried through when the tubeside
velocity seemed adequate to do this. This seemed to be the root cause.
6
The ‘docking’ and propagation of microbial biofilms is described as quite insidious, even overcoming
repellant electromotive charge between (negative) bacteria and (negative) surface, attributed in this
(biomedical) citation to millions of years of evolutionary adaptation.
308 10937790
This supplement to the .ppt : “AmerenUE Copper-Beryllium Tube Trial on Labadie Plant Unit 2”
outlines a proposed strategy for future work to further build on the operation record AmerenUE has
already established with a new type of heat exchanger tube material ( seam-welded copper-beryllium ).
Progress
As a contingency measure should through-wall failures of 439 stainless persist, in mid-2004 AmerenUE
pro-actively decided to install (26) tubes at our Labadie Plant Unit 2, composed (Figure 1) of a new type
of material (precipitation-hardened copper-beryllium)-- never before manufactured in a seam-welded
tube product form, nor ever placed in commercial use for an electric utility steam surface condenser.
The 0.035” thick strip stock was provided by BrushWellman, Inc. under a ladle analysis they developed to
be uniquely suited for the intended use (weldability and formability—i.e., tuberoll expansion of tube-ends
into condenser tubesheets- slide 8 ). BrushWellman sent the 56 HrB strip to a Sitindustrie facility in
Switzerland for seam-welding into a tube product, using their proprietary TIG open root, no filler WPS.
Also in 2004, the tubing composition, ultimate strength, etc. was submitted for review to the ASME
7
Boiler & Pressure Vessel Code (code case 2404) , under UNS designation C17510.
From there, this limited mill run quantity of new cu-be tubes were sent back to the U.S. to an age-
hardening (thermal processing) facility (slide 6), where they underwent 850F temperature soak at 5 hr.
hold time, full-furnace batch, conducted within an N2 inerting atmosphere / shipping cask to avoid
discoloration (slide 10).
After heat-treatment, the hardness of the cu-be tubes increased from 56 HrB to about 94Hrb, and
comparative micro-hardness (Vickers DPH, 100g) was also run vs. recognized TEMA tube materials
8
(slide 7). The excellent hardness ( while maintaining reasonable ductility ) was a major advancement
towards potentially substantially mitigating the shell-side wastage / copper-carryover concern which had
created the industry aversion to copper-based materials, and this was a motivating factor behind this trial /
field thermal performance evaluation.
Thermal Characteristics
The 26 trial copper-beryllium tubes have been in operation on AmerenUE Labadie Plant Unit 2 since
November 24, 2004. Individual tube-end exit temperatures were instrumented , and suffice is to say
(slide 16), that, when clean (i.e., 99.1% being attainable), temperatures were recorded to suggest a
formidable thermal conductivity for cu-be, as high as 188 Btu/hr-ft-F (compare 439 stainless at 12.7, a
225 value for pure copper , and 70 for the original Admiralty brass tubes). The cleanliness factors were
9
determined by modeling cu-be on Karta Technologies software as an equivalent length of 37.45’ long
copper-iron C194 ( the actual tube length at Labadie being 36’ ).
Unfortunately, longer term ( through May 31, 2005 ) the Missouri River has not been totally amicable
towards cu-be. Cleanliness factors have degraded to bulk condenser c.f.’s after 25-27 weeks (slide 18).
The thermal incentive to develop a means to sustain higher c.f.’s with cu-be ( target of 81- 89% ),
preferably without the need for the mechanical cleanings ( described earlier ), is substantial. Sensitivity
cases were run (slide 31) to identify that only a small cadre of 2200 clean cu-be tubes could outperform
10937790
309
over 15000 cu-be tubes running between 53 - 59% c.f. (slide 22), so the plausibility of use of an
improved ( ‘next generation’ ) cu-be tube product in mixed-metallurgy re-tubes certainly still offers
promise (not to mention in (larger) new-plant condenser tenders as well— 636mw and above, in which
the cost of biocide injection is more favorably offset by the avoided capital cost in tube quantity the
differential tubecost premiums apply to). Long-range (fully commercially deployed) target cost of such a
developmental (‘next generation’ ) cu-be tube is proposed as $5.96 / l.f.
Future Work – Next Generation Cu-Be
Is it possible to further surface re-engineer the cu-be material towards one that is inherently ‘biofilm
immune’—needing only periodic backflushes to sustain an exceptionally clean-running tube? Clearly,
Unilever’s biomedical research in the area of implanted cobalt-phthalocyanine transition metals to
10
produce plastic biomed. appliances (catheters, etc.) with a certain level of chemical activation is quite
relevant, insofar as intensifying the anti-microbial action of an extrinsically-applied biocide at the (root-
cause) referred to earlier ( i.e., at the biofilm –to-surface interface, or ‘docking’ sites ).
As the photo-negative (Figure 2) reveals of a cu-be sample (shown never even exposed to in-service
fouling), there are surface irregularities evident (tooling marks at Sitindustrie?) which are as-expected as a
normal part of the roll-forming. Nevertheless, are these scuff marks a sufficient discontinuity in surface
11
finish as to be pre-disposed ‘docking’ sites for biofilm adherence later in service? The question is not
easily dismissed.
Figure 2
The introduction of ‘precision surface engineering’ or ‘nano-layer deposition’ treatment to the tube
intrados as an intermediate manufacturing stage after roll forming and welding ( but prior to age-
hardening ) is an interesting possibility for on-going materials research.
12
In particular, ionic fusion of catalytic nano-layers for other applications (gas turbine blading) has been
performed successfully, and developing a variant ( in this case , co-phthalocyanine onto the i.d. of copper-
beryllium substrate prior to precipitation hardening ) for demonstration and thermal evaluation purposes
seems to be the next advancement.
310 10937790
References
1. Proceedings, September 2002 EPRI Condenser Technology Conference , San Antonio. Session 5:
Fouling Control –p. 5-12. SIDTEC Condenser Cleaning for Cooling Water Systems; R. Jones,
GE Betz. Trevose, PA.
2. February, 2004; ‘The World Market for Condensers for Steam or Other Vapor Power Units: A 2005
Global Trade Perspective. ICON Group International. http://www.mindbranch.com/products/R307-7340.html
3. Dohrer, Allan ; Yuba Heat Transfer.
4. AvestaPolarit Product Bulletin (Table 3.), ‘PRE for Stainless Steels (Pitting Resistance Equivalent)’
PRE number = %Cr + 3.3 (%Mo) + 16 ( %N)
5. May 12, 2005; Applied Specialites Inc., ‘Bacterialogical DNA Profile Analysis’ , water/slurry
sample #050505-A, for AmerenUE Labadie Plant.
6. Dunne, W. Michael, Jr. ‘Bacterial Adhesion- Seen Any Good Biofilms Lately?’; Department of
Pathology and Immunology and Department of Molecular Microbiology, Washington University
School of Medicine , and Microbiology Laboratory, Barnes-Jewish Hospital, St. Louis, Missouri .
Bacterial Adhesion: Seen Any Good Biofilms Lately? -- Dunne 15 (2): 155 -- Clinical Microbiology Reviews
7. Jawad, Maan PhD, PE President Global Engineering & Technology LLC

Cases of ASME Boiler And Pressure Vessel , Approval Date August 11, 2004: Case 2404
Copper-Beryllium Alloy (UNS C17510) for Code Construction, Section VIII, Div. 1
8. Laird Technologies website explaining precipitation hardening (see slide 9 of the AmerenUE
.ppt : “AmerenUE Copper-Beryllium Tube Trial on Labadie Plant Unit 2”).
9. Heat Exhanger Workstation –Condenser Application (‘HEW-CA’, v1.01 ); Karta Technologies

(copyright 2002—Electric Power Research Institute).
10. Wood, Paul; Jones, Martin; Bhakoo, Mohan; and Gilbert, Peter: ‘A Novel Strategy for Control of
Microbial Biofilms Through Generation of Biocide at the Biofilm-Surface Interface’ .
Microbiology Research Group, Dept. of Pharmacy, University of Manchester, Oxford Rd., Manchester
M13 9PL, England , and Unilever Research, Port Sunlight Laboratory, Wirral, L63 3JW, UK.
A Novel Strategy for Control of Microbial Biofilms through Generation of Biocide at the Biofilm-Surface Interface -- Wood et al. 62
(7): 2598 -- Applied and Environmental Microbiology
11. Vorburger, T. V. and Raja, J. ‘Internet-based Surface Metrology Algorithm Testing System ‘
(National Insitute of Standards & Technology – ‘NIST’, Gaithersburg, MD )
http://www.nist.gov/sigmaxi/Posters04/bui.html
12. Spadaccini, C.D., Center for Micro and Nano Technology , Lawrence Livermore National
Laboratory : ‘Diffusion of Platinum Films into Nickel-Based Substrates’ …
http://www.ionicfusion.com/files/pdf/Platinum Films.pdf, in collaboration with
Ionic Fusion Corporation…IONIC FUSION --- This Changes Everything!
10937790
311
Copper-Beryllium Condenser Tubing Trial
Ameren UE Labadie Plant unit 2
Presented by.. Glenn J. Tiffin , PE
. Co-Authors..
Bill Nielsen, Bob Kusner - BrushWellman,Inc.
Maan Jawad - Global Engineering &
Technology, LLC
( ASME Code Case 2404 )
Wayne Mueller- Supt. Quality, Ameren
Product forms- Why copper-beryllium?
Design Material Properties / Precipitation
Hardening
Development of a cu-be for injection systems?
Thermal Field Results
Economics (mixed-metallurgy re-tube)
312 10937790
Westinghouse Condensers on these 636mw Labadie units are 2-pressure,
1" x 0.049" Admiralty tubes ...
x 0.028" 439 grade stainless (existing re-tube)
x 0.035" copper-beryllium (this trial)
.
Several passes of carbon
steel scrapers were used
to clean all stainless test
tubes during a Labadie 2
. outage which ended
11 / 23 / 04
100% of all other tubes

( one shot, each ) as HP
c on
follows... den
ser
LP condenser- stainless
steel brushes LP c
ond
en ser
HP condenser- carbon
steel scrapers
No mechanical cleaning
.system, debris filter or
.biocide injection system .
.is in use on this unit.
# 6737 x 36' x 4 passes = 970,128 total l.f.
10937790
313
Pristine reference tube ('beta') plugs were pulled ...
We did this during

an 2 /18 /05 forced
unit outage on
Labadie 2 (in-
service was
11/23/04 ).
One concern
with the 2-Tube
method is
whether the
twisted-tip bead
also develops a
biofilm !
314 10937790
Where C17510 falls among other precipitation-hardened products...
turbine rotor wedges

for generator watch gears
blades windings
390
laptop
pressure switch docking
bellows, aircraft station
bushings contacts
98.1- 98.8
1.0- 1.4
0.15- 0.5
10937790
315
Composition of UNS C17510 seam-welded copper-beryllium tubes to ASME B&PV
Code case 2404…
Beryllium content was limited to 0.36% for reasons of maintaining maximum thermal
conductivity ...
CODE CASE OUR

RANGES HEAT
0.36%
0.04% compare 17310
1.57% at 1% co
0.02%
0.02%
0.02%
compare C194
97.93%
at 97% min.
The 30 tubes Ameren purchased from Brush-Wellman were seam welded in Switzerland
at a Sitindustrie mill and age-hardened at a facility in Tulsa, OK.
cu-be strip 56HrB tube

tube mill product Age hardening 223
supplier ( seam- facility (full Vickers
0.015mm welded, TIG furnace batch;
DPH
grain size open root, no
strip future plan for in-
filler) line @ tube mill )
316 10937790
Heat treatment t /c traces from age-hardening facility ...
10937790
317
We measured the hardness of several tube samples to verify the hardness of
copper-beryllium relative to recognized TEMA tube materials...
105 Vickers DPH microhardness, ex-works tube mill

100g test load, Diamond Pyramidal Hardness
Due to the excellent

hardness of cu-be,
improved wearability
under shell-side
wastage mechanisms
(i.e., reduced or
eliminated copper
carryover) was a major
motivating factor in the
trial.
318 10937790
copper-beryllium retains
excellent ductility and fatigue strength
despite it's exceptional hardness --i.e.,
it's hardness did not inhibit as-installed
%-deformations under tube-end roller
expansion to be run into our grooved
muntz-metal tubeholes higher than
mock-up qualification testblock values
(8% vs. 5%, with no evidence of
ligament distress).
Copper-beryllium exhibits 'composite' material properties

- Why ... ?
10937790
319
This is an explanation from the website on precipitation hardening...
320 10937790
Shown below are two of the as-
delivered copper beryllium tubes
Our 30 tubes were delivered as shown >> which were rejected for excessive
( heat treatment was carried out in an N2 tube-end out-of- straightness
inerting cask for purposes of eliminating tolerance . We rejected 4 of 30 tubes
charring to the surface appearance ) which did not respond to a rattler
during insertion
Our daqs system for 2-tube field temperatures ( heat treatment cask / shipping
(our waterboxes are in background) container is shown )...
10937790
321
Labadie Condensers were re-tubed in 1999 from Admiralty to 439 gr. stainless and developed 30
instances of through-wall tube-pitting (source of cooling water: Missouri River)
Although we've had tubeleaks, an economic case cannot be made (later slides) that there are significant lost
mwhs due to silica problems during the economic recovery criteria period (first 5 yrs.) after a stainless re-tube.
The only lost mwhs are due to biofouling ( i.e., unit turndown to keep b.p. < 5" HGa ).
After installing a
ball-cleaning system
on Labadie 4 in
2003,
26 copper-beryllium
tubes were also
installed on Labadie
2 in Oct.-Dec. '04 for
evaluation,
chiefly as a
contingency plan
should failures
persist.
We are also starting to
seriously consider
biocide injection.
.
322 10937790
Stainless is a general ( but perhaps controllable ) problem as a heat exchanger
tube selection in organic exposure due to the loss in thermal conductivity
caused by the high degree of alloying needed to build in a high PREN
( "pitting resistance equivalent " number...)
As shown, 439 stainless, with a PREN of 17 , has the lowest PREN value....
PREN = %Cr + 3.3 %Mo + 16 %N
14
439
fyi, several tube manufacturers make 12
exotic specialty grades
' k ' (btu/ h ft F )

grades of stainless with the highest 10 29-4C
PRENs available, but note how the 8
thermal conductivity drops off with 6
increasing PREN... 4
0
0 10 20 30 40 50 60
PREN
10937790
323
But even pure copper can develop a biofilm leading to
performance degradation or tubeleak ...
source: copper.org (seawater exposure)...
25 -27
weeks
324 10937790
R&D STEERING PROPOSAL FOR TUBE Mfr'g. INDUSTRY
consistent with EPRI promulgating

thinking beyond our own industry
in this case, the biomed. & semiconductor (nano-layer deposition) industries
Long-term development of a cu-be tube product form with electro-deposition
of a cobalt phthalocyanine thin film might someday eliminate biofouling...
http://aem.asm.org/cgi/content/abstract/62/7/2598
10937790
325
Design basis used in equivalent tube-wall calculation
C194
C194
326 10937790
Piece-wise continuous characterization of the cuprous materials series, by similitude to
equivalent length on HEW-CA ...
The expected value of 'kcu-be' makes

sense where it falls considering
linearly extrapolating beyond that to
100% copper tube (known TEMA 'k' of
225) @ a finite lequiv.
10937790
327
HISTOGRAM OF 8 DATASETS, EACH OF 2hr. DURATION
Cleanliness factors shown were determined on a Karta v1.01 HEW-CA
software simulation ( using the equivalent thermal length of cu-be
modeled as 0.035" wall copper-iron C194, 37.45' long ).
118ºF 4/1/05 4/29/05 5/16/05 5/31/05 6/28/05 7/6/05 7/21/05 8/22/05
53.2 ºF 58.3 ºF 68.6 ºF 74.5 ºF 79.0 ºF 81.2 ºF 85.4 ºF 84.3 ºF
river river river river river river river river
3:00 - 5:00 pm 3:00 - 5:00 pm 3:00 - 5:00 pm 3:00 - 5:00 pm 3:00 - 5:00 pm 3:00 - 5:00 pm
628mw 626mw 624mw 624mw 616mw 627mw
Max. capability of unit was dropping off too low

to make any further representative comparisons
No calc.- use of
silver solder on
twisted beads
may have
69.4% introduced
67.6 % error *.
83.1 % 65.2 %
74.8 %
78.8 %
53.0 %
74.1 % 79.4% 49.6 %
{
91.1% 75.8%
into the test program

HP 77.3 %
no calc
85.4 F
99.1%
7/14/05, Top FWH

82.6 % 36.6 %
taken oos 7/19/05

Signals restored*
Cond. 76.5 % 83.1 % 56.9 %
79.0 F
76.0 %
71.7 %
74.5 F
{
73.7 %
LP 68.9 %
Cond. 67.2 %
61.7 % 68.6F
68ºF
328 10937790
c.f.'s referenced to baseline bulk conditions...
* economic ' base case' (later slide) is based on these values of bulk c.f.'s. Note how much higher these minimums
would have to be if macro bias to the LP inner were to be present...
10937790
329
Premise for future 2200-tube cu-be demonstration w. biocide injection
see ' case iii ' economics, later slide
Effect of macro-fouled bias to LP-inner pass...
addntl. est. net cost of662,000

$436k
330 10937790
LP inner macro bias was measured on another Labadie unit ...
10937790
331
Nominal Base case, ' ' ( lowest minimum c.f.'s in 3 of 4 passes for no load reduction, if
it weren't for the occurrence of LP inner macro-bias )
base case
332 10937790
Sensitivity: 15,237 (uncladded) cu-be tubes, but running no cleaner than base case ' '
case i -a
nearly 57% of
condenser
copper-
beryllium
10937790
333
Heat rate benefit of -0.329"Hga, or -0.8% heat rate decrease (-80 btu/kwh) @ $160k annual
savings . . .
case i -b
334 10937790
Economic recovery of premium of +$1.76/ ft. (uncladded cu-be) v. $2.20/ft super-
ferritic stainless, 15237 cu-be tubes
case i economics
10937790
335
Biocide injection added to outside pass (only) of base case ' ' to improve c.f.'s
( and load ) to those shown
case ii-a
336 10937790
Heat rate benefit of -0.644"Hga, or -1.2% heat rate decrease (-120 btu/kwh) @ $240k
annual savings . . .
case ii-b
10937790
337
Incremental economics of one-side biocide injection ...
case ii economics
338 10937790
Same as case ii, but 2200 (uncladded) cu-be tubes substituted as shown ( biocide injection system
enables -13,037 tubes or $436K net avoided capital cost v. "case i " at $1.76/ ft. (uncladded
(uncladded)) premium v. super-
super-
ferritic s.s.
s.s. bulk re-
re-tube stock* )
*super-ferritic s.s. base cost
i.e., tubecost assumed $2.20 / ft.
savings, less
$390k
capital cost
of biocide
injection
case iii-a
10937790
339
case iii-b
340 10937790
Incremental economics of tubecost premium component of +$1.76/ ft. (uncladded cu-be),
v. $2.20/ft super-ferritic stainless (compare to ' case i ' )
case iii economics
Although assumptions are involved in the

would-be performance of an injection
system, one of the major findings of this
study was quantifying the extent to which a
smaller cadre of cu-be tubes (on larger
condensers) have better economic return
than larger numbers of cu-be tubes
(without injection) that don't run as clean .
10937790
341
Same as case iii, but use of cu-be (cladded) tubes running exceptionally clean
(reduces $436k to $278K net avoided capital cost at $1.99/ ft. (cladded) premium v. uncladded cu-be )
case iv-a
342 10937790
case iv-b
10937790
343
Incremental economic recovery for the target cladded cu-be tubecost component for a
2200-tube cu-be demonstration with biocide injection ( i.e., $5.95 / l.f. total tubecost)
case iv economics
344 10937790
Conclusions......
Copper-beryllium has been seam welded in tube product form for the first time and is in trial use in a
. condenser alongside 439 stainless tubes exposed to the same biofouling. Plans are in place to reduce reject
. rate due to tube-end distortion during heat-treat by in-line age hardening / roll-straightening for quantity-
. production mill runs. Qualification of a no-filler WPS, perhaps under ASME SB-543 is also planned
. (code case 2404 is currently only approved for plate, using filler).
Field temperature data / HEW-CA similitude indicate a ' k ' for cu-be possibly as high as 188 Btu / H ft F. .
New cu-be tubes degrade to condenser bulk c.f. in 25-27 weeks, along with 'double-scraped' stainless tubes.
636mw units: (' case iv ') - economic return is favorable ( < 2 yrs., at +52.6mw and -185 btu/ kwh) for a
. 2200-tube, $687k mixed-metallurgy cu-be demonstration, incorporated in a planned re-tube, assuming $297k
. in (co-phthalocyanine cladded) cu-be tube cost premium + $390k cost of a (one-.side) biocide injection system.
Smaller (278mw) units: ( ' case i ' ) would probably be preferred over (' case ii, iii or iv ') - no biocide
. injection, just under 57% of the tube population cu-be, or 7060 uncladded cu-be tubes, at a $433k premium.
Target cu-be tube costs of $3.96 /ft (uncladded)

. and $5.95/ft. (cladded) were established vs.
. $2.20/ ft. for 'super-ferritic' stainless, creating
. an industry incentive to consider development
. of co-phthalocyanine catalytic thin films +
. (electro-deposited over the tubeside of cu-be
. substrate - during strip manufacture?-) that might
. enhance peroxide effectiveness / reduce feed rate.
Removal of a cu-be tube specimen for condition assessment is planned (o.d. tubeseal area, etc.).
10937790
345
Grateful Acknowledgements
Maan Jawad- Global Engineering & Technology, LLC
Bill Nielsen- Brush-Wellman
Wayne Mueller- Supt. Quality, Ameren
David Fox- Plant manager, Labadie
Mark Litzinger, Tony Balestreri, Greg Bolte, Jim Vaughn, Brian Griffen-- Labadie Engineering
Keith Fann & Mike Bray - Labadie IT
Doug Buck & crews- Conco
Forrest Truemper- Matrikon
Robin Goatey- Ameren Business Process Infrastructure Specialist
Dale Karq- Santee Cooper nuclear
Mike Catapano- PowerFect
John Eimer, Ramon MiraFlores, Brian Jones & crews- Ameren GCMS
Labadie electricians & technicians shops & Labadie Operations station crews
Lori Kaltenbronn, Gary Rogles, Scott McCormack - Ameren GTSS Lab Svcs.
Shannon Abel-- Ameren Callaway nuclear
Peter Viccaro- Omega Engineering
Ivan Franson (ret)- Allegheny Ludlum
Todd Meyer- Ameren Generation Projects
EIC Solutions
AvestaPolarit
Karta Technologies
Allan Dohrer- Yuba Heat Transfer
Laird Technologies
Jack Maurer- Nickel Development Institute (NiDI)
346 10937790
GRAY WATER COOLING IN SURFACE CONDENSERS AND HEAT
EXCHANGERS
TREND OR ABERRATION
Dennis J. Schumerth
Manager - Titanium Products
VALTIMET, Inc.
14281 Franklin Avenue
Tustin, CA 92780 USA
10937790
347
GRAY WATER COOLING IN SURFACE CONDENSERS AND HEAT
EXCHANGERS
TREND OR ABERRATION
Dennis J. Schumerth
VALTIMET, Inc.
ABSTRACT:
In recent years, concern over the continued use of limited fresh water supplies or similarly,
cooling towers and their essential makeup, high maintenance and associated chemical
treatment requirements has spawned a crafty, yet dramatic change in powerplant surface
condenser and heat exchanger cooling. The paradigm shift away from the established and
typical toward the unconventional has produced an innovative and non-traditional cooling water
source for surface condensers and heat exchangers. Clearly, gray water-cooling has come of
age.
Pundits suggest water shortages will increase the amount of water reuse (Chart 1) in the US
from a current estimated 1.7 billion gallons to an estimated 12 billion gallons by the year 2015 2.
Chart 1
Projected Water Reuse
15
Billions gal/day
10
0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Given this dramatic prediction, water reuse, or the reclamation and treatment of impaired or
gray water will be driven by and emerge as market drivers joined at the hip by emerging effluent
discharge standards. Without a clear understanding of the legislative and political landscape,
regulative complexities that deal with this type of cooling water could conceivably lead to an
unattractive, environmental legacy.
348 10937790
Having duly noted the “trend or aberration” dilemma, this paper will further identify efforts by the
municipal wastewater treatment plants to economically process a usable product. We will
investigate the impact of ancillary add-on costs absorbed by the electric utility such as
secondary filtration and examine an actual case study involving the extensive use of gray water.
Finally, the paper will evaluate new operational conditions, emerging new corrosion issues with
suggested abatement, metallurgical changes, pollution considerations, maintenance issues and
other mechanisms which have forced utilities to develop innovative solutions when employing
impaired cooling water sources for the main surface condenser and other heat exchangers.
OVERVIEW
By definition, gray water is cooling water where all or part of the flow stream is made up of either
partially or fully treated sewage effluent. As you can well imagine, the use of sewage effluent
provokes a plethora of new issues. They are led by the voluminous unknowns that flow from
society to the sewage treatment plant and the economics of processing and transporting this
impaired water from the municipal host to the ultimate user. The application of this relatively
new cooling medium suggests the potential impact of this “water” on plant metallurgy, chemical
treatment requirements, corrosion abatement and other physical plant system needs, can
become a blueprint for both the speculative and the unproven.
BACKGROUND
Of the 24,000 municipal wastewater treatment plants in the U.S., it is estimated that only about
1,500 employ water reuse facilities. Indeed, more glaring is the fact that only 6% of the total
municipal wastewater volume is presently reused. This percentage is even less when applied to
power generation facilities. Economic, legislative and logistical impediments to wholesale
expansion of water reuse appear to be the high cost associated with medium transport,
biological nutrient removal, macro and mircofiltration, ultraviolet disinfecting and corrosion
abatement activities.
In addition, the “relative” abundant supply of fresh water, be it destined as make-up or once
thru, poses even greater challenges to the increased use of impaired water. Even though
legislative action, albeit confusing, is currently underway to curtail the use of this “fresh” water,
current regulative issues, high transport cost, interruptible shortages, wastewater disposal and
inconsistent quality all contribute to any dramatic increase in the use of impaired water. In stark
contrast, the relatively stable and predictable cost of fresh water undermines, in many cases,
the unpredictability of sourcing to impaired water. Given this operational and economic
conundrum, it should be noted that a number of utilities and utility consortiums have
successfully made the transition from fresh to impaired. This has been accomplished by
maintaining a successful economic return – both within the operating utility itself and the
community at large.
Costs
Chart 2 2 below identifies the comparative raw cost of water worldwide. You will note the United
States enjoys a relatively low cost when compared to other locations. One could speculate that
this low, first cost poses economic roadblocks to the usage enlargement of impaired water. In
many areas of the country, this is a truism. The first or raw cost of the water is not however,
10937790
349
Chart 2
Water Cost
WATER COST Water Cost

COUNTRY
($/1,000 gal) ($/m3)
Germany $6.70 $1.78

Puerto Rico $5.00 $1.32
Netherlands $4.31 $1.14
Italy $2.75 $0.73
Finland $2.43 $0.64
Southern California $2.27 $0.60
United States * $2.06 $0.54
Canada $1.42 $0.38
Note: Costs west of the Mississippi River can be well above the national average
the final cost of treated water. The below graphic (Chart 3) will identify the add-on costs to treat
a variety of waters using both the “conventional” (chemical) and MF/RO (micro filtration/reverse
osmosis) processes.
Chart 3
Treated Water Costs
BASE CONVENTIONAL MF/RO

COST U.S. TREATMENT TREATMENT
$2.06/1000 gal $2.84/1000 gal $2.68/1000 gal
The Process
Each day, U.S. industries consume 25 billion gallons of water all while generating about 20
billion gallons of wastewater. Furthermore, each day, thermoelectric plants in the U.S. consume
186 billion gallons of water. Given the voluminous flows and declining resources, gray or
impaired water, as an alternate cooling medium, has emerged as a trendy option.
Should sewage effluent be considered as the cooling medium, a first or initial treatment typically
takes place at a municipal sewage facility 13. Here, raw effluent is processed to (1), physically
separate solids from liquids and (2), purify the liquid.
Preliminary Treatment: Solids, such as wood, rags and plastic are removed by
screens. This debris is washed, dried and removed for safe disposal. Grit and sand are
similarly removed.
350 10937790
Graphic 1
Typical Sewage Treatment Process
Primary Treatment: Remaining solids are separated from the liquid using large,
settlement tanks. The settled solids, referred to as sludge, are further treated for use as
fertilizers.
Secondary Treatment: Biological or percolating filters break down organic material and
purify the liquid. The process can be speeded up using aerating tanks. Further
separation to isolate sludge is also required during this treatment phase.
At this point, the sewage or gray water is suitable for transport to the user facility. If
further or tertiary treatment is required, final “polishing” may be required before the water
is returned to the environment.
Phoenix - Tolleson, Arizona Wastewater Treatment Facility
10937790
351
Chart 4
Typical Condenser Cooling Water Analysis 7
Parameter Effluent (mg/l) Parameter Effluent (mg/l)

Fecal Coliform 13 Sodium 139
Fecal Coliform 7 Hardness 320
BOD 5.5 Calcium 71.6
TSS 5 Magnesium 34.4
COD 47.3 Alkalinity 97.2
Nitrate+ite 21 Carbonate 0
Ammonia <0.1 Bicarbonate 119
TKN 2.66 Chloride 65.2
T Phosphorus 2.23 Fluoride 0.67
Potassium 10.8 Sulfate 312
pH 7.49 TDS 730
Heavy Metals Heavy Metals
Effluent (mg/l) Effluent (mg/l)
(total) (total)
Aluminum 0.15 Magnesium 34
Antimony <0.001 Mercury <0.0002
Arsenic <0.001 Molybdenum 0.003
Barium <0.1 Nickel <0.01
Beryllium <0.001 Selenium <0.005
Boron 0.3 Silicon 7.1
Cadmium <0.001 Silver <0.001
Calcium 73 Strontium 0.7
Chromium 0.002 Thallium <0.001
Cobalt <0.001 Tin <0.1
Copper 0.01 Uranium 0.002
Iron 0.1 Vanadium <0.001
Lead <0.001 Zinc 0.07
Once the effluent arrives at the powerplant site, it undergoes a series of further treatments.
Initially, trickling filters are employed to reduce ammonia and alkalinity. Additional multi-phase,
biochemical treatment processes typically employ clarifiers where phosphates, magnesium,
silica and some calcium are removed. A second stage removes much of the calcium-carbonate
(CaCO3) using several chemical treatment options. Calcium-carbonate, if not addressed, can
be a significant source of scale buildup and corrosion concern. Sulfuric acid may be added at
this point to reduce pH and chlorine is added to control biological growth. A final gravity filtration
will remove remaining suspended solids.
Palo Verde Wastewater Treatment Plant
352 10937790
At this point the treatment is complete and the gray water is transferred to storage reservoirs
and used as tower makeup. In other cases, the treated effluent can be used directly as the
main cooling water.
Palo Verde Plant Site & Storage Reservoir
Corrosion Discussion
Calcium-carbonate
The transformation of sewage effluent to gray water – water suitable for use in a powerplant
surface condenser produces unusual issues that deserve special attention. The first, which was
noted previously, is the identification and reduction of phosphates or calcium-carbonate
(CaCO3). The data show 8 that calcium-carbonate can initiate formation after only 1.5 cycles of
concentration. A typical water analysis of gray water would suggest CaCO3 levels range in the
68 ppm (City of Amarillo, TX) to 71 ppm (Raton, NM). Xcel Energy 8 has noted that CaCO3
becomes a problem at levels around 15 ppm. Higher cycles of concentration of the tower or the
effluent itself will clearly exceed the 15 ppm threshold and precipitate in the form of deposits on
the condenser tube ID surface opening up the potential to underdeposit pitting in susceptible
tube materials. To minimize the fouling buildup, several solutions are can be employed.
Chart 5
Soda ash &

Ferric Continuous on-line Oligomers
Lime carbon dioxide
Sulfate cleaning (scale inhibitors)
gas
In many locations, selection of the lime dosing proved the most economical. Others have
selected the soda ash/carbon dioxide treatment. Evaluation of scale inhibitors classified as
oligioers show promise as they are chlorine resistant 1. An on-line cleaning system will also
10937790
353
prove beneficial if already in place. However, the capitol expenditure of a new unit may prove
prohibitive.
Chlorine
A highly effective measure to prevent biocidal growth in all treatment areas typically includes the
use of chlorine. This is especially true in pipelines and areas of the treatment that are highly
susceptible to the spread of fecal coliform staff infection. Some utilities will use gaseous
chlorine sparingly and have eliminated all forms of chlorine shock due to reasons of safety,
regulatory and public relations moving instead to bleach/bromide combinations. The use of
chlorine in gray water applications is particularly troubling. If the cooling water contains amines
or ammonia, chloroamines are formed which consume chlorine increasing the dosage amount
to achieve the desire effectiveness.
Manganese
Manganese can also contribute significantly to corrosion concern. Recent research articles
identified what they refer to as one of the most interesting and insidious corrosion issues
relating to underdeposit pitting in impaired water cooling systems. It has to do with corrosion
that is caused by manganese oxide - a phenomena that causes severe pitting on the tubes. The
corrosion mechanism is not completely understood, however, it appears that soluble
manganese precipitates as manganese dioxide on the condenser tube surface. Indeed, the
manganese may be naturally occurring in river or lake water, or in sediments. If sediments
become anaerobic, the manganese in them can solubilize. The soluble manganese
subsequently oxidizes and precipitates as manganese dioxide on condenser tubes. The
austenitic family of stainless steel appears to be particularly susceptible to the phenomenon of
manganese induced, underdeposit pitting.
One possible explanation for the corrosion is that oxidizing biocides - such as chlorine - oxidize
the manganese oxide to soluble permanganate. This destroys the passive layer on stainless
steel and creates cathodic and anodic areas that generate severe pitting. Some researchers
also theorize 5 that biofilms themselves can concentrate manganese oxide. When the biofilm
contains iron and manganese-oxidizing bacteria then can create manganese-oxide deposits on
the tubing. These deposits may work in conjunction with sulfate-reducing bacteria, creating
corrosion cells. It would appear this statement is counterproductive in that the use of chorine
may actually cause corrosion issues. However, since the chlorine must be used in the effluent
treatment process, there appears little choice in the matter. This can be a double-edged sword
should the cooling water contain amines or ammonia – clearly present in gray water. In this
environment, chloroamines are formed which consume chlorine and thus increase the amount
of chlorine required to produce the desired results.
An additional problem with manganese is that it induces pitting by changing the potential of the
exposed material. In the case of surface condensers, titanium is immune to this type of attack
because it has such a very high pitting potential (on the order of +10V) 6. SS, on the other
hand, has a pitting potential very close to its rest potential (less than +1V) and can be
susceptible to pitting attack when oxidizing compounds are present that raise the potential.
Because all stainless materials are susceptible to their own PRE critical pitting temperature
number 11, care must be exercised in the proper material selection when manganese-oxide
conditions are present or suspected.
354 10937790
There also appears to be a connection or overlap between impaired or wastewater reuse for
cooling and MIC/manganese attack. The case history noted in the reference data 5, identified
manganese-induced corrosion of stainless steel piping at the clarified outlet of sewage
treatment plants. Higher cycles, waste, minewater reuse, hyper-chlorinization and drought
conditions can exacerbate the problem.
MIC
Invariably, Microbiologically Influenced Corrosion (MIC) must be addressed when employing

impaired or gray water. The bacteria present will predictably, place susceptible materials in
harms way. The susceptibility of stainless steels to MIC is well documented 6,9,10. In particular,
304/304L and 316/316L are at risk. Indeed, batch culture tests indicate that all alloys examined
at the time (316L, 904L, Al-6X, 254 SMO & 625) are susceptible to MIC attack 10. Later tests
suggest the “N” grade of AL-6X exhibited good resistance to MIC 6. Considerable testing by the
Naval Research Lab 6,11 suggests titanium is immune to MIC – even at elevated temperatures
(55 – 70oC).
Floaters and Sinkers
Effluent water quality can vary a great deal from city to city – from source to source. Plastic
materials (floaters) can accompany the effluent water floating on the top of clarifiers potentially
plugging heat exchanger equipment. Suspended solids and debris (sinkers) tends to form
sludge in the cooling tower basin. Initially, chlorine was used to reduce the biological fouling
identified as sulfate reducing bacterial.
However, heavy chlorine dosing can cause damage to system metallurgy – particularly the
brass family of condenser and heat exchanger tubing. Other methods may be employed as a
result of this damage potential. Similarly, high concentrations of ammonia will cause harm to
copper bearing materials. High BOD also tends to exacerbate the problem.
A Case Study
Over the past several years, an increasing number of new generating facilities have employed
the use of gray water in some form of cooling – either as tower makeup or, in some cases,
direct, once-thru cooling. Delta Energy, Millennium Power, Bosque, The City of Lakeland,
Florida, Londenderry and PSE&G are just a few of the subscribers to this latest trend. Indeed,
PSE&G – both Bergen & Linden stations replaced their new 316 stainless condenser tubing with
titanium principally because of the highly corrosive nature of the cooling medium. Xcel Energy
and SWEPCO have employed effluent cooling in some form at their Nichols and Jones
Generating stations for almost 40 years
Historically, the recent spate of impaired water usage has been for small to mid-size generating
units. The Arizona Public Service – Palo Verde Generating Station (APS-PV) is the glaring
exception in terms of sheer size and historical precedence. Palo Verde is a three-unit, PWR
facility generating a total of 3,875 MW and provides electric power to 4 million people in the
Southwest. The station has been in operation since 1986 and uses gray water exclusively for
cooling. See Chart 5 for plant statistics.
10937790
355
Cooling Water
Raw sewage, received from the greater Phoenix area, is initially treated at the Tolleston, AZ
Municipal Sewage plant before transport approximately 45 miles via a 96” diameter line to Palo
Verde. Additional treatment is completed and purified water is pumped to the on-site storage
reservoir for use in the closed loop condenser/tower circuit. A nearby CCGT unit – Redhawk
also uses a partial flow of the effluent.
Palo Verde Storage Reservoir & Cooling Towers
Cooling Towers
Three (3), mechanical forced-draft cooling towers service each generating unit. Corrosion of
exposed rebar and spalling of the concrete into a gelatinous substance has occurred over the
operational years due to the continuous wet/dry cycling. Chlorinating has been used to
successfully combat the algae growth. The tower operates at 25 concentration cycles resulting
in salinity ≥ than seawater. Once this concentration is reached, the water is discharged to
evaporation ponds.
Chart 6
DISTANCE SEWAGE PUMPED 45 MILES
DIAMATER OF PIPE 96”
SEWAGE PLANT FLOW 58 MILLION GALLONS/DAY
APS-PV WATER RECLAMINATION CAP 90 MILLION GALLONS/DAY
STORAGE RESERVOIR 670 MISSION GALLONS 80 ACRES
TOWER EVAPORATION (ave) 14,000 GAL/MIN/UNIT

TOWER BLOWDOWN (ave) 865 Gal/min/unit
EVAPORATION POND 250 ACRES – 2 PONDS
TOWER CONCENTRATION CYCLES ≥ 25
356 10937790
Condenser
A Marley, 3-pressure, 3-shell surface condenser was field erected and tubed (see on-site
erection photo) at the APS-PV site. In classic multi-pressure, variable tube length configuration,
the CIRH20 is series connected in a parallel path (allowing bundle isolation) from the LP to IP to
HP shell (see condenser schematic). The cycle is completed with shell C discharging to the
cooling tower. Because of the corrodents present in the cooling water, the tube material was
changed from stainless steel to titanium. The tube bundles were later staked to prevent the
onset of damaging vibration due to the excessively large support plate spacing. The tubesheets
are Al. Bronze with mechanically expanded tube joints. The tubesheets are coated at the inlet
end but not at the discharge. All 12 water boxes are coated.
It is of keen interest to examine the overall performance and integrity of this condenser given its
considerable service life, after-the-fact tube material selection, galvanically dissimilar tubesheet
material, tube-tubesheet joint configuration, coating philosophy and above all, the aggressive
water chemistry. Let us consider the following after nearly 20 years of service life.
On-site Surface Condenser Erection (1 of 6 shells)
1. No titanium tubing corrosion has taken place.
2. The integrity of the tube-to-tubesheet joint appears to remain viable.
3. Coating the inlet tubesheet has apparently halted some initial erosion of the Al.
Bronze material. The initial erosion may have been galvanically induced.
4. Some fatigue failures of the titanium tubes were attributed to excessive support
plate spacing. Staking successfully addressed this issue.
5. Several tube failures resulted from poor design of the cold water discharge
spargers.
10937790
357
6. Some minor steam erosion has been detected on the tube OD at the top of the
bundle.
7. Mechanical scrapers are used to keep the tubes clean. Little to no ID buildup
has been observed using this cleaning method.
Conclusion
The remarkable and increasing use of sewage effluent to cool powerplant component systems
including main surface condensers and ancillary heat exchangers has been successfully
demonstrated at many locations where fresh water is unavailable, not usable or too costly.
Considering the rather undesirable source, impaired or gray water, given proper treatment, has
emerged as an economically viable and highly sustainable resource. As a result, it becomes
clear that water reuse will increase dramatically over the next 10 years notwithstanding the
invasion of regulatory complexities that could derail this continued growth pattern.
The paper followed the transformation process from effluent to potable water taking on the
nuances of multi-phase treatments, attendant corrosion mitigation and metallurgical “red flags”.
User experience suggested chemical treatments and material selections should be implemented
based solely on good engineering practice. Engineers need to take a highly pragmatic view
when considering material options, limitations and selection within the operating environment.
Considerable dialog was spent on the manganese and chlorine issues – issues that can
dramatically impact the operational competency of the system. Finally, a brief case history
study of the APS - Palo Verde experience demonstrated that these practices can be
successfully implemented on a long-term basis within the operating environment. Proven
technologies and good engineering practices, not myopic speculation must be employed when
operating in such a volatile system.
Graphic 2
CONDENSER SCHEMATIC
CT
Turbine C/L
Shell A
Shell B
Shell C
CIRH20
358 10937790
References:
1. Power – September 2002 – “Untangling the Complexities of Cooling Water Chemistry”,

David Daniels, Contributing Editor.
2. Water Executive Article – September/October, 2004 - U.S. Water Reuse Expected to
Grow Substantially During the Next Decade”, Ultra Pure Water Conference, April 15 &
16, 2004. Mike Lenley & J. Seibert
3. The grateful contributions of Jer Chin Shih & Frank Francuzik, Arizona Public Service,
Palo Verde Nuclear Generating Station
4. Corrosion, 2003 - Wayne H. Dickinson, Ph.D. Various excerpts & opinions
5. Corrosion 2003 – Paper 03563
6. Corrosion93 – An Evaluation of Titanium Exposed to Thermophilic & Marine Biofilms,
Brenda Little, Patricia Wagner & Richard Ray – Naval Research Laboratory, Stennis
Space Center
Corrosion93 – An Experimental Evaluation of Titanium’s Resistance to
Microbiologically Influenced Corrosion, Brenda Little, Patricia Wagner & Richard Ray –
Naval Research Laboratory, Stennis Space Center
7. City of Raton, NM
8. ASME Chemistry Committee, April, 2004 – Sewage Effluent for Cooling, Xcel
Energy’s Two-Score Experience, Bernie Wieck
9. NACE Task Group T-5A-28, Corrosion90 – Microbiologically Influenced Corrosion of
Austenitic Stainless Steel Weldments, Susan Watkins Borenstein, PG&E
10. NACE Task Group T-5A-28, Corrosion90 – Ranking Alloys for Susceptibility to MIC –
A Preliminary Report on High-Mo Alloys, P.J.B. Scott, Joyce Goldie & Michael Davies,
CARIAD Consultants, Toronto Ontario
11. Industrial Water Treatment Vol. 28 No 4 & No 6, 1996 – Corrosion Problems and
Countermeasures in MSF Desalination Plant Using Titanium Tube, Fukuzuka et al
12. Vallourec Information on Corrosion
13. City of Tolleson, AZ Municipal Wastewater Treatment Plant
10937790
359
GRAY WATER COOLING IN SURFACE
CONDENSERS AND HEAT EXCHANGERS
TREND OR ABERRATION
Principal Author: Dennis J. Schumerth

VALTIMET, Inc.
Phone 714-573-1000 FAX 714-426-0001 Cell 714-393-1013
E-mail schumert@AOL.com
Webb Site: Valtimet.com

August 30 – September 1, 2005
Catamaran Resort Hotel
San Diego, California
360 10937790
T yp ical Im p aired C o n d en ser C o oling W ater An alysis
P a ra m e te r E fflu e n t (m g /l) P a ra m e te r E fflu e n t (m g/l)

F ecal C o lifo rm 13 S o d iu m 139
F ecal C o lifo rm 7 H ard n ess 320
BOD 5.5 C alciu m 71.6
T SS 5 M ag n esiu m 34.4
COD 47.3 Alk alin ity 97.2
N itrate+ ite 21 C arb o n ate 0
Am m o n ia < 0.1 Bic arb o n ate 119
TKN 2.66 C h lo rid e 65.2
T P h o sp h o ru s 2.23 F lu o rid e 0.67
P o tassiu m 10.8 S u lfate 312
pH 7.49 TDS 730
He a vy M e ta ls He a vy M e ta ls
E fflu e n t (m g /l) E fflu e n t (m g /l)
(to ta l) (to ta l)
Alu m in u m 0.15 M ag n esiu m 34
An tim o n y < 0.001 M ercu ry < 0.0002
Arsen ic < 0.001 M o lyb d en u m 0.003
Ba riu m < 0.1 N ickel < 0.01
Be rylliu m < 0.001 Selen iu m < 0.005
Bo ro n 0.3 S ilico n 7.1
C ad m iu m < 0.001 S ilv er < 0.001
C alciu m 73 S tro n tiu m 0.7
C h ro m iu m 0.002 T h alliu m < 0.001
C o b alt < 0.001 T in < 0.1
C o p p er 0.01 U ran iu m 0.002
Iro n 0.1 V an ad iu m < 0.001
L ead < 0.001 Z in c 0.07
10937790
361
Projected Water
Projected Water Reuse
Reuse
15
Billions gal/day
10
0
03
05
09
15
01
07
11
13
20
20
20
20
20
20
20
20
362 10937790
Object of the Paper
Examine the Process to Achieve a Usable Product

Evaluate Operational Conditions
Identify Emerging Corrosion Issues
Suggest Appropriate Abatement Procedures
Identify Maintenance Issues
Present Actual Case Study
10937790
363
Gray Water or “Impaired Water”
Definition: Flow Stream = Fully or Partially Treated Sewage Effluent
Usage: < 6% Wastewater is Reused
The Good: Readily Available in Untreated Form
The Bad: Economic & legislative impediments + ROI
The Ugly: Potentially Higher Cost

Medium Transport
Biological Nutrient Removal
Marco & Micro Filtration
UV Disinfecting
Corrosion Abatement
364 10937790
Wa t e r C o s t
WATER COST Water Cost
COUNTRY
($/1,000 gal) ($/m3)
Germany $6.70 $1.78

Puerto Rico $5.00 $1.32
Netherlands $4.31 $1.14
Italy $2.75 $0.73
Finland $2.43 $0.64
Southern California $2.27 $0.60
United States $2.06 $0.54
Canada $1.42 $0.38
10937790
365
Treated Water Cost
BASE CONVENTIONAL MF/RO

COST U.S. TREATMENT TREATMENT
$2.06/1000 gal $2.84/1000 gal $2.68/1000 gal
366 10937790
Typical Sewage Treatment
Process
What is the Object ?
Separate Solids
From Liquids…..
or…….
“Floaters from Sinkers”
10937790
367
Typical Sewage
Treatment Process
What is the Object ?
Purify The Liquid
368 10937790
Preliminary Treatment
Screens Remove Solid Debris
Primary Treatment
Settlement Tanks Separate Liquid
From Solids
Secondary Treatment
Percolating Filters Break Down
Organic Matter & Isolate Sludge
Transport to Host or Final Polishing
10937790
369
Primary Treatment
From Solids
Secondary Treatment
370 10937790
Primary Treatment
From Solids
Secondary Treatment
10937790
371
Primary Treatment
From Solids
Secondary Treatment
372 10937790
Tolleson Arizona Wastewater
Treatment Facility
Aeration Tank Pumping Station
10937790
373
Treatment Plant
to the
Power Plant
374 10937790
Palo Verde Wastewater
Treatment Plant
Additional Treatment
•Trickling Filters
Reduce ammonia and alkalinity
10937790
375
Treatment Plant
•Clarifiers - Stage 1
Removes phosphates, magnesium
silica and calcium
376 10937790
Treatment Plant
silica and calcium
Removes calcium carbonate
10937790
377
Treatment Plant
silica and calcium
•Sulfuric acid controls pH
378 10937790
Treatment Plant
silica and calcium
•Chlorine Controls biological growth
10937790
379
Treatment Plant
silica and calcium
•Chlorine Controls biological growth
•Gravity Filters remove suspended solids
Gray Water Treatment is Now

Complete!
380 10937790
Corrosion Considerations
1. Calcium -Carbonate (CaCO3)

Threshold Limits
Range 15 - 71 ppm can initiate problems
Concentration Limits
1.5 to 25 cycles can exceed threshold limits
Remedial Action Options
Soda ash &

Ferric Continuous on-line Oligomers
Lime carbon dioxide
Sulfate cleaning (scale inhibitors)
gas
10937790
381
2. Biocidal Growth (fecal coliform/staff)

Treatment Options
Chlorine
Bleach/bromide Combinations
Caveat
Grey water may form chloroamines
which will consume chlorine
requiring dosage increases
382 10937790
3. Manganese
Manganese Oxide = underdeposit pitting
Chlorine treatment may exacerbate the problem
Grey water may compound the problem
Austenitic stainless is susceptible

Proper material selection
10937790
383
4. MIC
Gray water bacteria places suspect materials in harms way!
Highest susceptableilty
304/304L & 316/316L
Other suspect materials
904L, AL-6X, 254 SMO, 625 (NACE, Naval Research Lab)
Resistant Materials
AL-6XN, UNS 44660, 29-4C
Immune Materials
Gr..2 Titanium (Naval Research Lab)

Proper Material Selection
384 10937790
5. Sinkers & Floaters (Debris)

Intelligent debris housekeeping
Sludge removal

Clarifiers free of solid materials
10937790
385
A Case Study
Arizona Public Service

Palo Verde Nuclear Generating Stations
3-unit PWR
3875 MWe
Cooling Source: 100% Treated Sewage Effluent
386 10937790
Palo Verde Cooling Towers
Palo Verde Evaporation Pond
Palo Verde Storage Reservoir
10937790
387
Plant Statistics
DISTANCE SEWAGE PUMPED 45 MILES
DIAMATER OF PIPE 96”
SEWAGE PLANT FLOW 58 MILLION GALLONS/DAY
APS-PV WATER RECLAMINATION CAP 90 MILLION GALLONS/DAY
STORAGE RESERVOIR 670 MISSION GALLONS 80 ACRES
TOWER EVAPORATION (ave) 14,000 GAL/MIN/UNIT
TOWER BLOWDOWN (ave) 865 Gal/min/unit
EVAPORATION POND 250 ACRES – 2 PONDS
TOWER CONCENTRATION CYCLES = 25
388 10937790
After Twenty Years of Service…..
10937790
389
2. The integrity of the tube-to-tubesheet joint appears to
remain viable.
390 10937790
remain viable.
3. Coating the inlet tubesheet has apparently halted
some initial erosion of the Al. Bronze material. The
initial erosion may have been galvanically induced.
10937790
391
remain viable.
4. Some fatigue failures of the titanium tubes were
attributed to excessive support plate spacing.
Staking successfully addressed this issue.
392 10937790
remain viable.
5. Several tube failures resulted from poor design of the
cold water discharge spargers.
10937790
393
remain viable.
6. Some minor steam erosion has been detected on the
tube OD at the top of the bundle.
394 10937790
remain viable.
6. Some minor steam erosion has been detected on the
tube OD at the top of the bundle.
7. Mechanical scrapers are used to keep the tubes clean.
Little to no ID buildup has been observed using this
cleaning method.
10937790
395
Conclusion
1. Sewage Effluent Cooling is on the rise
CT
2. Proper treatment is vital
3. Regulatory issues do not appear to be a factor.
4. Metallurgical ‘Red Flags”
5.Turbine
ProperC/L chemical treatments & material selections
Shell A
are imperative Shell B
6. Palo Verde case
Shell Chistory
CIRH20
396 10937790
References:
1. Power – September 2002 – “Untangling the Complexities of Cooling Water Chemistry”,

David Daniels, Contributing Editor.
2. Water Executive Article – September/October, 2004 - U.S. Water Reuse Expected to
Grow Substantially During the Next Decade”, Ultra Pure Water Conference, April 15 &
16, 2004. Mike Lenley & J. Seibert
3. The grateful contributions of Jer Chin Shih & Frank Francuzik, Arizona Public Service,
Palo Verde Nuclear Generating Station
4. Corrosion, 2003 - Wayne H. Dickinson, Ph.D. Various excerpts & opinions
5. Corrosion 2003 – Paper 03563
6. Corrosion93 – An Evaluation of Titanium Exposed to Thermophilic & Marine Biofilms,
Brenda Little, Patricia Wagner & Richard Ray – Naval Research Laboratory, Stennis
Space Center
Corrosion93 – An Experimental Evaluation of Titanium’s Resistance to
Microbiologically Influenced Corrosion, Brenda Little, Patricia Wagner & Richard Ray –
Naval Research Laboratory, Stennis Space Center
7. City of Raton, NM
8. ASME Chemistry Committee, April, 2004 – Sewage Effluent for Cooling, Xcel
Energy’s Two-Score Experience, Bernie Wieck
9. NACE Task Group T-5A-28, Corrosion90 – Microbiologically Influenced Corrosion of
Austenitic Stainless Steel Weldments, Susan Watkins Borenstein, PG&E
10. NACE Task Group T-5A-28, Corrosion90 – Ranking Alloys for Susceptibility to MIC –
A Preliminary Report on High-Mo Alloys, P.J.B. Scott, Joyce Goldie & Michael Davies,
CARIAD Consultants, Toronto Ontario
11. Industrial Water Treatment Vol. 28 No 4 & No 6, 1996 – Corrosion Problems and
Countermeasures in MSF Desalination Plant Using Titanium Tube, Fukuzuka et al
12. Vallourec Information on Corrosion
13. City of Tolleson, AZ Municipal Wastewater Treatment Plant
10937790
397
10937790
6
SESSION 6: MAINTENANCE TECHNOLOGY II
10937790
399
10937790
MAIN CONDENSER
TUBE PLUGGING
Mark Meltzer
PSEG Nuclear LLC
August 31, 2005
10937790
401
CONSIDERATIONS
• Tube Mapping
• Plugging Criteria
• Tube vs. Tubesheet
• Tube Preparation
• Plugging Issues
• Plugs Available
• Plug Issues
• Conclusions
402 10937790
TUBE MAPS
• Software set up to show an exact
replica of your tubesheet(s) – inlet and
outlet
• Can show much more than just plugged
tubes
– Eddy current history
– On line and available to everyone
10937790
403
PLUGGING CRITERIA
• 50% through wall or greater
• All restrictions
• All non-quantifiable indications
• Known issues
• Eliminate all unknowns from the

condenser
404 10937790
TUBE PREPARATION
• Thoroughly clean and dry tube
• Consider wire brushing to remove
debris and / or organic material
• Tube should not be deformed or
blocked
• Make sure you have the right plug for
the tube
10937790
405
TUBE vs. TUBESHEET
• Significantly different application
• Harsher environment
• Different size hole to fill – diameter and
length
• Different material???
• May be grooved
• Poor tubesheet plugging typically has
much higher impact on plant
406 10937790
PLUGGING ISSUES
• Deformed tube
• Tubesheet coating in the way
• Blockage / dent in tube
• Periphery tube
10937790
407
INTERNET VENDORS
• CONCO Systems, Inc.
– http://www.concosystems.com (800-345-3476)
• The Conklin Sherman Co., Inc.
– http://www.conklin-sherman.com (203-881-0190)
• Expansion Seal Technologies Group, Inc.
– http://www.expansionseal.com (800-355-7044)
• HEPCO Heat Exchanger Products Corp.
– http://www.hepcoplugs.com (800-472-8484)
• John R. Robinson, Inc.
– http://www.johnrrobinsoninc.com (800-726-1026)
408 10937790
INTERNET VENDORS (cont)
• Specialty Tool & Supply, Inc.

– http://www.spectool.com (800-360-8665)
• JNT Technical Services, Inc.
– http://www.torq-n-seal.com (201-641-2130)
• Thermal Solutions, Inc.
– http://www.tubeandpipetools.com (610-379-5446)
• TUPROS, Inc.
– http://www.tupros.com (937-325-5044)
• USA Industries, Inc.
– http://www.usaindustries.com (800-456-8721)
10937790
409
BASIC TYPES OF PLUGS
• Pin
• Pin and Collar
• Expanding Rubber
• Relaxing Rubber
• Expanding Metal
410 10937790
CONCO Systems, Inc.
High Confidence Pin and Collar
Fibre Tapered
Expanding
Rubber
10937790
411
The Conklin Sherman Co., Inc.
Relaxing Rubber
412 10937790
Expansion Seal
Technologies Group, Inc.
Perma Plugs Small Ram Package
10937790
413
HEPCO Heat Exchanger
Products Corp.
Expanding Rubber
414 10937790
John R. Robinson, Inc.
Expanding Rubber
Pin
Pin and Collar
Relaxing Rubber
10937790
415
Specialty Tool & Supply, Inc.
Expanding Metal
Pin and Collar
416 10937790
JNT Technical Services, Inc.
Expanding Metal Relaxing Rubber
10937790
417
Thermal Solutions, Inc.
Pin
Pin and Collar
Expanding Metal
418 10937790
TUPROS, Inc.
Pin
10937790
419
USA Industries, Inc.
Pin Pin and Collar
Pro-Plug Snap-It
Expanding Rubber
420 10937790
PLUG ISSUES
• Improper installation
• Corrosion
• Design
• Normal wear
• Environment
10937790
421
IMPROPER INSTALLATION
Plug not
properly torqued
in
422 10937790
IMPROPER INSTALLATION
Plugs over extended
10937790
423
CORROSION
Hardware
corroded
away
424 10937790
DESIGN
Difficulty of
installation and
maintenance
10937790
425
NORMAL WEAR /
ENVIRONMENT
How long is the

plug designed
to last in the
environment
where it will be
used
426 10937790
CONCLUSIONS
• Not convinced there is a single best tube plug
• Plugging on-line vs. outage
• Tube vs. tubesheet – different needs
• Sizing and installation are key
• Verify periodically
• How long will rubber last???
• Cover / seal with epoxy
10937790
427
10937790
NEW DEVELOPMENTS IN CONDENSER TUBE COATINGS
Michael J. Horn
James E. Mitchell
Plastocor, Inc
100 Research Road
Hingham, MA 02043
Abstract
Premature failure of steam condenser and balance of plant heat exchanger tubing is generally
due to pitting. Mitigation of tube pitting may involve improved cleanliness, chemical treatment, or
less commonly, the coating of tube interiors. Tube coating has been uncommon due to various
issues including: field application quality control, cost, and the possible negative impact on heat
transfer. A new method for coating tubes has been introduced which addresses these
drawbacks. This method fills any tube pits with coating material while also applying a very thin
layer to the tube wall. Here the pits are remediated and the tube is protected from further pitting.
The thin layer of coating also promises to help with tube cleanliness by reducing fouling rates.
The conventional methods for coating tubes are described and the new method of tube coating is
outlined and discussed.
Introduction
The tubes used in steam condensers and other balance of plant heat exchangers are selected
such that they will provide many years (decades) of reliable service. In the case of heat
exchangers, reliable service generally consists of being leak free with acceptable heat transfer
properties. Tube materials are specified with regard to expected water chemistry, cleanliness,
and cost. However, due to unforeseen circumstances or to operational anomalies, tubes may
begin to form pits, which if not mitigated or otherwise controlled, can result in tube leaks. In most
cases, the tube must then be identified and plugged. Significant costs are incurred locating and
plugging a tube, all because of a single pit. Identifying the mechanisms for pit formation and
understanding how to control these processes is an important aspect of the maintenance of
steam condensers and other heat exchangers.
Common Causes for Tube Pitting
Microbiologically Influenced Corrosion (MIC) - Because of the lack of light and complex nutrients,
the predominant form of microorganisms, which induce pitting in tubes, are bacteria which draw
nutrients from the cooling water. Many types of bacteria exist whose byproducts or chemical
presence can create a corrosion cell in condenser and heat exchanger tubing. This corrosion cell
results in a pit, which enlarges the habitat for the bacteria, which results in greater corrosion
activity. Very complex pit structures can be formed by these bacterial colonies. The depth and
complexity of these MIC pits can make the eradication of these microorganisms extremely
difficult.
Under deposit corrosion – various electrochemical and/or microbiological activities can occur
under tube deposits that set up corrosion cells.
Under film corrosion – Depending on water chemistry, films of various types can form on the tube
walls. Certain kinds of film such as sulfide or manganese have galvanic potentials, which can
differ from the base metal which can result in pitting if they are disturbed.
10937790
429
NEW DEVELOPMENTS IN CONDENSER TUBE COATINGS Page 2 of 8
2005 EPRI Condenser Technology Seminar and Conference
Conventional Responses and Remediations
While the response to a pitting situation is dependent upon the situational specifics, one or more
of the following actions are generally followed:
- Improved cleanliness – The underlying cause of many pitting phenomena is a lack of

cleanliness. The removal of tube deposits and the installation of better tube cleaning
methods may be sufficient to stop tubes from further deterioration.
- Chemical treatment – While general chemical treatments are usually a means to better
cleanliness (scale reduction/corrosion control), other treatments utilizing biocides or
oxidizing agents can be used to control biological activity which is initiating the pitting.
Biostats are used to inhibit the growth of microorganisms and represent a milder form of
control, perhaps following the use of a biocide.
- Eddy current testing and preventative plugging. Due to the depth and/or morphology of
the pits, the conventional tools may not be effective in stopping existing pits from
continuing to grow. In this case it may be necessary to identify those tubes with deep
pits (usually via eddy current testing) and to preventatively plug those tubes. This is only
effective as long as the underlying cause for the pitting has been identified and corrected.
- Retube – In severe cases, where the pitting can not be effectively be controlled, or an
unacceptable amount of tubes have been lost, the tubes must be replaced. In this case
care must be taken to choose a more appropriate tube material or take steps to insure
that the conditions which caused the original pitting do not reoccur.
Tube coating as an alternative
In the process industry, tube coatings are a common approach to corrosion control. In refineries
and chemical plants, carbon steel tubing is often specified and then coated to protect it from
deterioration. This approach is cost effective because of the abundance of heat exchanger
bundles which are routinely removed for cleaning or replacement which allows for the proper
application and curing of tube coatings in a shop setting.
In the power industry, the use of coatings on tubesheets is quite common whereby corrosion, and
erosion effects are eliminated and tube-to-tubesheet joint integrity and strength are restored. The
use of coatings on the ends of heat exchanger and condenser tubes is also common in the power
industry. Here the effects of inlet-end or outlet-end erosion/corrosion are mitigated and
permanently controlled by applying a feathered 6 to 9 mil coating to the tube-end interiors where
turbulence often causes selective wear patterns.
Since 1984, Plastocor, Inc. has applied its’ coatings to over one thousand condenser tubesheets
and seven million condenser tube-ends.
430 10937790
Full-length tube coatings for power plant heat exchange equipment, however, have not been
generally accepted. There are several reasons for this including:
- Power plant heat exchange bundles are less abundant than process industry bundles
and are not routinely removed for service; main steam condensers are fixed structures
and cannot be removed for off-line shop maintenance. Therefore, the primary method for
controlling tube corrosion has been the selection of the proper metal alloy, which can
resist the anticipated service environment for the long-term. Heat transfer considerations
are calculated based on expected cleanliness and do not anticipate the use of coatings.
- Full-length tube coatings are difficult to apply properly in the field both because of the
difficulty in properly preparing the tube surface, and of the difficulty in applying a uniform
coat to thousands of tubes at a time. Proper curing of solvent-based coatings in the field
has also been an issue given the confining nature of a long tube, which does not allow a
proper escape route for the solvent. This leads to an uncertain performance outcome for
coated tubes.
- Concerns about the heat transfer effects of the coating. Coatings are inherently poor
thermal conductors and can have a detrimental effect on heat transfer depending on their
thickness.
- The cost advantage for properly coating tubes has not been sufficiently compelling, given
the uncertain outcome, when compared with the certainty of the more costly retubing
option.
Traditional Tube Coating Methods
Traditionally, tube coatings have been applied either by flow coating or by spray. With the flow
coating method, the tube (or the whole bundle) is flooded with coating and then drained and
cured. Curing is often done at elevated temperature in an autoclave, and multiple coats are
usually required to build up the specified coating thickness. This process must be done in a shop
where the tube bundle can be manipulated. With the spray method, a 360° spray nozzle must be
centered and sent down the length of each tube and retracted at a consistent speed with a
precisely controlled coating flow rate. The spray nozzle may be connected to a hose or lance and
controlled by various mechanisms. Care must be taken to be sure that the apparatus and spray
nozzle are operating properly for the duration of the project. Solvented coatings are generally
used with these systems and care must be taken to insure that the solvent has escaped and that
the coating is cured properly. Multiple coats may be required to achieve the desired coverage.
New Tube Coating Method
In addressing the specific issues associated with pitting of power plant condenser and heat
exchanger tubing, a new approach to tube coating has been devised. The root of this approach
is a change of emphasis from that of conventional coating.
With conventional tube coatings the goal is to:
- Apply a uniform film in one or more coats over the entire length of the tube.
- Have the thickness of the film be sufficient to confer prophylactic protection to the
tube surface.
10937790
431
This approach assumes that every part of the tube needs that same amount of protection which is
typical of why coatings are used, that is to protect an inherently vulnerable substrate. The
problem with this approach for alloy heat exchanger and condenser tubing which has pitted is that
the thickness of the coating film must be such that it remediates (fills/covers) the pits. If the
coating is applied thinly, then there may be an inadequate amount of coating in the pit. This
condition is illustrated in Figure 1.
Figure 1: Tube with pit and inadequate amount of coating to remediate pit
Conversely, if the coating is applied at a sufficient quantity to adequately fill the pit, the coating
runs the risk of being too thick on the rest of the tube wall. This excess of coating is likely to
seriously impair heat transfer of the tube. This condition is illustrated in Figure 2.
Figure 2: Tube with pit and excessive amount of coating on tube wall
With the new approach to tube coating described in this paper, the goals are to:
- Target and fill (remediate) the existing pits
- Apply a uniform but minimal film in one coat over the unpitted surface of the tube so
as to minimize heat transfer loss but confer some protection. This condition is shown
in Figure 3.
432 10937790
Figure 3: Tube with pit which has been filled and with minimal coating on tube wall
With this approach it is acknowledged that the pitting was unanticipated and can be controlled by
better maintenance practices, such as improved cleanliness. Here by repairing the existing pits
and preventing further development, the integrity of the tube can be preserved and its life
extended. It is further noted that because the tube is of a resistant alloy (not carbon steel), those
areas, which are not pitted, do not need a thick film of coating to protect them. A minimal film will
provide sufficient protection while not overly affecting heat transfer. In addition, this coating film
can be engineered to provide additional benefits such as fouling resistance.
Specifics of the New Method
Cleaning of the tube
Prior to the application of the coating, the tube must be thoroughly cleaned. Generally this
consists of three steps: rough cleaning, decontamination, and final cleaning.
Rough Cleaning: This consists of removing the preponderance of scale or other fouling from the
tube so that it can be effectively decontaminated. Rough cleaning can be accomplished by
means of hydrolazing, scrapers, chemical cleaning, brushes, sponge ball system, or any other
method that will effectively remove the major deposits.
Decontamination: Generally this means being certain that the tube is free of chlorides or other
contaminants which could interfere with the performance (adhesion) of the coating. This can be
accomplished by flushing the tubes with demineralized, low chloride potable water, or water
conditioned with a chloride neutralizer.
Final Cleaning: This consists of the final cleaning which removes all residues and moisture from
the tube in preparation for coating. This is generally done via abrasive blasting.
The goal of these steps is not only to clean the tube wall but to clean out the pits as well. The
final abrasive blasting has been found to be effective in scouring pits if performed properly from
each end of the tube.
10937790
433
Coating of the Tube
Instead of using a spray head to disperse the coating throughout the tube, it was decided to
develop a pigging technology whereby a “slug” of coating is introduced into one end of the tube
which is to be coated. A specially designed and sized coating pig is inserted into the tube and
propelled down its length with compressed air, or drawn down the tube with a rod or line. The pig
pushes the “slug” of coating along the tube causing it to enter and fill the pits as it travels through
the tube.
In this case, the pig is designed to remove, or squeegee off, as much coating as possible as it
travels through the tube. This is accomplished by giving the pig a sharp edge at its trailing end.
The result is that the pits remain filled with coating while the tube interior in general is left with a
very thin film of coating. Coating thickness on the tube wall is commonly less than a mil. A
patent is pending for this method of tube lining.
The elements of this “squeegee pig” approach are shown in Figure 4.
Figure 4: Schematic (exaggerated) of Pitted Tube and Squeegee Pig in Tube Coating Process
Characteristics of the Tube Coating
While a variety of coating materials can be used with the above described method, there are
several essential characteristics:
x 100% solids – A coating is said to be 100% solids if it does not contain a solvent which
must evaporate for the film to cure. The confined nature of a long heat exchanger tube
makes it difficult for a solvent to escape. In addition, it is even more difficult for the
solvent to escape from a pit that has been filled with coating. A solvented coating will
also shrink far more appreciably than an unsolvented one which is problematic for filling
pits.
434 10937790
Characteristics of the Tube Coating (cont.)
x Low Viscosity -- Because the coating, does not contain solvents to control its viscosity, it
must be compounded so that it is sufficiently low in viscosity to: readily wet out the tube
wall, flow into small pit structures, and be easily squeegeed from the tube wall for thin
wholesale application.
x Durable: The coating must be compounded so that it can withstand the normal operating
conditions of the condenser/heat exchanger. This includes both day to day operations as
well as periodic tube cleaning methods.
x Thermally Conductive: While coatings are generally not considered to be thermally

conductive, they can be made to be more conductive by selecting conductive fillers
and/or applying a very thin layer.
For use with this methodology, specialized epoxy compounds have been found to have the best
combination of the above characteristics.
In addition to the above essential characteristics the following beneficial characteristics can be
engineered into the coating:
x Deterrence of Microbiological Activity – The coating may be compounded with a biocide

and/or biostatic agent to enhance the control or elimination of deleterious microbiological
activity in the condenser/exchanger tube.
x Deterrence of Macro Fouling: The coating may be compounded so as to resist fouling by

being sufficiently slick (low surface energy). These properties may also enhance fluid
flow through the tubes which could boost heat transfer by allowing higher flow velocities.
Coating Thickness and Heat Transfer
Heat transfer in condenser tubes is commonly broken down into five categories:
x Fluid film resistance on the waterside of tubes. 39%

x Oxidation, scale, and other fouling resistance on the waterside of tubes. 33%
x Tube wall metal resistance. 2%
x Oxidation and other fouling resistance on the shell (steam) side. 8%
x Fluid film resistance on the shell (steam) side. 18%
1
The percentage each contributes is drawn from the work of Lustenador & Staub
The actual extent to which each of these elements affects overall heat transfer varies with tubing
material, operation parameters, and water quality. Shell side effects are not affected by coatings.
For the purpose of evaluation, coatings are best viewed as becoming part of the oxidation, scale,
and other fouling resistance. Evaluating the effect of the coating on the heat transfer can then be
described in terms of its effect on the cleanliness factor of the tube.
10937790
435
In order to get a preliminary indication of the coating’s effect on tube cleanliness, two new 7/8”
OD, 304 Stainless, 22 BWG tubes were coated and compared to an uncoated tube section. The
thickness of the coating was the variable for this test. The cleanliness tests were performed by
Bridger Scientific, Sagamore Beach, MA who has developed a test method which simulates the
thermal loads experienced by condenser tubes. The test parameters were setup to replicate
typical full load design conditions as specified by the Heat Exchanger Institute. The results of this
test are:
x Tube 1: no coating; cleanliness level = 100%
x Tube 2: coating thickness of approximately .6 mils; cleanliness level = 92.7%
x Tube 3: coating thickness of approximately 1.6 mils; cleanliness level = 69.2%
While these results are preliminary, it suggests that a thinly applied coating (under 1 mil) could be
acceptable for use on a typical steam condenser. This conclusion is based on typical design
calculations which assume an operational cleanliness factor of 80 to 90 percent. In addition, if
the coating is found to have a beneficial effect in regards to further fouling, these numbers look
even better.
In contrast, it is seen that a thicker coating has a much more meaningful impact on heat transfer
which, depending on operating conditions, may not be acceptable.
Conclusion
By focusing on the need to reverse the progression of pits in condenser tubes, and to minimize
the heat transfer impact, a new method for applying coatings to tube interiors has been created.
This method uses a patent pending “squeegee pig” to push and distribute a coating through the
tube which then fills any existing pits and applies a thin layer of coating to the unpitted surfaces of
the tube interior. It is envisioned that this method will provide an efficacious and economical tool
which can be used to prolong the life of power plant heat exchanger and condenser tubing.
Areas for further research include:
x Coating design to improve heat transfer characteristics and fouling resistance

x Application methods to improve efficiency of the operation
x Surface preparation methods
x Inspection methods
x Testing to determine durability and longevity of the tube lining
1. E.L. Lustenador and E.L. Staub, “Development Contribution to Compact Condenser

Design,” International Nickel Power Conference, May 1964
436 10937790
CONDENSER RENOVATION AND THE APPLICATION OF EPOXY COATINGS ON
BRASS TUBES
Jean F. Brunel
Electricite de France
Abstract
The renovation of 32 of EDF Nuclear Power Plant condenser’s, has been going on since 1990 and will
continue until about 2013. The 20 year-old condensers, initially fitted with brass tubes, must be
refurbished because of tube thickness reduction, caused by erosion and abrasion from the cooling
water (raw water). Old tubes are replaced either by titanium, stainless steel or brass tubes. The choice
of material depends on the technical and environmental constraints of each site. Carbon steel tube
sheets are coated with the Plastocor® thick layer epoxy cladding system, the inner surfaces of the
waterboxes are coated with a Plastocor® thin film epoxy, the inlet of the brass tubes are coated with 3
layers of a similar Plastocor thin film epoxy and an experimental epoxy coating was put on the entire
length of certain tubes.
Renovation Methods
The renovation of EDF condensers is made either by modular bundle exchange or re-tubing.
The particular method is chosen depending on the economical balance of each solution. Renovation
costs are compared to increased earnings, for the remaining plant life, by the level of electricity
production expected after renovation. Thus, modular exchange is chosen for condensers whose
performances can be improved, whereas re-tubing is the solution for optimized condensers.
Within this same frame of technical and economical analysis, EDF also examined methods to improve
the life of the new brass bundles, as well as methods to extend condenser tube life for units near
renovation time or final shutdown. Therefore EDF decided to experiment with epoxy coatings on the
inner surfaces of the brass tubes. These coatings were put just on the inlets or by coating the entire
length of the tube.
Epoxy Coating Selection
All of the materials used on the tubesheets, waterboxes, tube inlets and tubes are 100% solids, zero
VOC (volatile organic content) epoxy systems. The basic chemistry used on these systems has over
thirty years of service history.
The full length tube lining material is an extremely low viscosity, flexible and abrasion resistant
material, which is put on at a thickness of between 100-175 microns. This formulation allows for
resistance to tube vibrations and poor surface conditions.
The inlet coating is a low viscosity, ceramic filled system. It is put on the ends of the tubes to protect
them from wear or repair them when damaged. The material is put on in three consecutive coats of
about 50-75 microns each at a length of between 10-30 centimeters. Each coat is put on slightly
farther down the tube so a smooth transition is made onto the tube wall. This procedure eliminates
problems with step erosion which is common when using either metal or plastic inserts. The inlet
coating can be used either with or without the full length tube lining.
10937790
437
Condensers Renovation and Epoxy Coating Tests
In FESSENHEIM 2, a sample of 375 brass tubes was coated in 2002, including 150 full length tube-
lining, 150 tube-lining plus inlet coating and 75 tube inlet only coating.
Coating operations were performed according to the schedule of 18 days and without any technical
problem. The initial measurement of the epoxy coating highlighted that the thickness of the layers
were complying with the specification, 100 µm for tube lining and 150 µm for tube inlet.
Visual inspection and thickness measurement will be performed during each unit shutdown until 2006.
The first results of those inspections, after 23,000 hours of operation, did not show any significant loss
of thickness.
Discussion of Results: Tube Inlet a Curative and Preventive Solution
As CRUAS condensers had reached their technical end of life in 2000, they were included in the EDF
general renovation program and scheduled for renovation between 2000 and 2004.
On this site, brass tubes were particularly affected with inlet tube erosion, leading to frequent tube
breaks and raw water intake. Those events used to induce a significant loss of production, due to load
reductions for water leak detection, and expenses as far as the intervention of specific labor teams was
needed.
In order to limit those consequences and to improve reliability until renovation time, inlets of the most
sensitive ½ of the condenser have been epoxy coated. The challenge for this condenser was not only
to implement an efficient solution, but also to coat very thin inlets.
The condenser operated for one year and the result is satisfying, as we estimate that 39 water leaks
have been avoided, meaning that production savings equal to 10 times the investment cost of the
solution.
This method for condenser protection is now frequently used by EDF, as every new brass bundle is
fitted with inlet epoxy coating, CRUAS 1 to 4, DAMPIERRE 2. The inlet tube protection is always
implemented along with a thick epoxy cladding on the tubesheet, and a thin coating of the waterboxes’
inner surfaces.
Thus, EDF expects a significant lifetime improvement for the renovated condensers.
Discussion of Results: Tube Lining a Balance to be Found
In spite of a technical established feasibility, full length tube lining isn’t yet identified as an
economical balanced solution for brass bundles of units near final shutdown.
This optimization should be re-considered for each condenser according to the evolution of risks
induced in terms of availability, and according to the evolution of phasing between renovation needs
and final shutdown.
Conclusions : EDF Condensers Lifetime and Availability Improved: the main reasons
for tube coating.
As a conclusion, EDF is now able to compensate one of the weakest point of brass tubes and can
therefore aim not only a significant lifetime improvement for renovated condensers whatever the
choice of tube material, but also a better reliability of condensers which are near renovation time or
final shutdown.
438 10937790
HEAT EXCHANGER TUBE STRUCTURAL REPAIR SLEEVES
Bruce W Schafer
Framatome ANP, an AREVA and Siemens company
155 Mill Ridge Road
Lynchburg, VA 24502
(434) 832-3360
bruce.schafer@framatome-anp.com
James H Batton
Duke Energy
Oconee Nuclear Station
7800 Rochester Highway
Seneca, SC 29672
(864) 885-3893
jhbatton@duke-energy.com
ABSTRACT
Defective heat exchanger tubes can be repaired using a variety of techniques that do not involve
the cost and time of component retubing or replacement. Various structural sleeves that are
designed to bridge the defective area of the tube, have been developed. These sleeves can be
used to repair damage done by excessive tube vibration, stress corrosion cracking, pitting, or
erosion.
Four sleeve designs are discussed within this paper. These are tube end stiffening, tubesheet,
deep bundle, and full-length sleeves. Each has a different application and each repairs a unique
tube degradation condition. However, all are structural repairs, allowing the sleeve to become
the pressure boundary in the event of complete tube failure. The sleeve installation process
typically includes a combination of roll expansion, full-length hydraulic expansion, or end
hydraulic expansion.
By installing a sleeve, the majority of the tube’s heat transfer and flow capacity is maintained.
This can effectively extend the operating life of the component. In addition, sleeve installation
will allow previously plugged tubes to be returned to service to reclaim lost heat transfer and
flow area.
INTRODUCTION
It is not uncommon for heat exchanger tube failures to occur during power plant operation.
From a sampling of industry data, tube failures have been shown to cause between 31% to 87%
(depending on the data source) of the events related to feedwater heaters (1). Failure of
condenser tubes, leading to the ingress of service water into the condensate system, is an all too
frequent occurrence. Furthermore, tube damage in one component has led to operability issues
with other heat exchangers that are used to cool components critical to the operability of power
plants (such as steam generators or boilers).
10937790
439
Since many of these failures are related to tubing, the replacement of the entire heat exchanger is
an expensive, as well as a schedule and manpower intensive, option.
Since the 1970s, structural tube sleeving has been used to allow damaged tubes to remain in
service as an alternative to heat exchanger replacement. The sleeves, which consist of a smaller
piece of tube placed inside of the parent tube, can be fixed in place by various means, such as
roll, explosive, or hydraulic expansion or explosive welding. The sleeve can become the new
pressure boundary for the damaged tube if the sleeve is properly designed and installed. The
sleeve will allow the tube to remain in service for an extended period of time, economically
extending the life of the heat exchanger. The decision to perform sleeving also can be made with
short notice as opposed to replacement, possibly allowing repairs to be performed during the
same outage that the damage is noted.
While not all of the field experience listed within this paper is for condenser applications, each
type of sleeve listed has been used in service water systems and could be used in condensers.
Detailed information is provided on the Oconee-1 Reactor Building Cooling Unit sleeve since it
is a new application of sleeving technology.
SLEEVE REPAIR – FACTORS TO CONSIDER

There are various factors to consider when deciding whether to repair the heat exchanger tubes
or to replace the entire heat exchanger. These are:
Available Budget
Typically, the cost of performing a substantial heat exchanger repair (consisting of plug removal,
tube inspection, and sleeving) is less than 10% of the cost of replacing the unit. Because of the
lower cost, the payback time on the repair option is much shorter than for replacement.
Plan Scheduled Replacement

If the heat exchanger is critical to plant operation (either from a safety, efficiency, or power
production standpoint) or is resulting in costly forced outages, it may be possible to justify a
temporary sleeve repair to the unit in the near-term and a scheduled component replacement
when a longer outage can be planned.
Plug Types and Quantities

If there are a large number of tube plugs to remove, or if they are difficult to remove (explosive
or welded), then the cost to repair the heat exchanger will increase, and the scheduled time
needed on-site may not fit within the outage window. If it appears that tube repair may be
possible, it may be worthwhile to use removable plugs until a certain quantity of tubes are
removed from service. At that point the plugs would be removed and sleeves installed, thereby
minimizing the overall maintenance cost.
Extent of Tube Damage

Tube repair may be appropriate if damage is limited to a certain area of the tube, which would
allow the use of a short repair sleeve. If the damage is over a significant portion of the tube, it is
possible to install a longer sleeve (up to the full length of the tube) to ensure that all tube defects
440 10937790
are repaired. However, it would not be possible to install a sleeve if a large portion of the tube
had damage and there was inadequate clearance for a long sleeve at the tube end.
Overall Condition of Heat Exchanger

The condition of the shell side components, such as the impingement plates, tube supports, end
plates, and other structural members, should be in good shape if a long-term repair is being
planned. An evaluation also should be made of the shell thickness in areas that are prone to shell
erosion/corrosion. If the tube repair is only a short-term fix, to allow component operation until
a replacement heat exchanger can be installed, the condition of the shell side is not as critical.
Remaining Plant Life

If the only problem with the heat exchanger is in one section of the tube, and the expected run
time on the unit is relatively short, it would be advantageous to repair rather than replace the heat
exchanger since it will be very difficult to pay back the cost for replacement over the remaining
plant life.
Outage Time Available for Heat Exchanger Repair

The outage time required to repair a heat exchanger, even when tube and shell side inspections
are performed, is typically much less than for replacement. In addition, very few, if any, plant
modifications need to be made to make the repairs. This allows other work to be performed in
the vicinity of the heat exchanger.
Site Support Requirements

Usually, there are no shell or head modifications required for tube side repair since all work can
typically be performed through the manways and pass partition plates. Less repair equipment is
required, resulting in less space needed in the area of the heat exchanger for setup and storage.
Time to Prepare for Heat Exchanger Repair

The time required to prepare for tube repair is much less than for replacement (2-8 weeks
compared with 18 months), allowing a decision on repair to be made just before, or even during,
an outage.
Disposal Cost of Old Component

At nuclear plants, the added cost for the disposal of radioactively contaminated heat exchangers
must be taken into account. Before disposal, there is the cost of surveying the heat exchangers
for release and, if contamination is found, they must either be decontaminated or disposed of as
radioactive waste. Tube repairs can eliminate these costs.
SLEEVING REPAIR OPTIONS

Various structural sleeve repair options exist. These sleeves are designed to form a new pressure
boundary within the tube and can withstand the same operating and design conditions as the
original tube, even if the parent tube has a 100% through wall indication. Both ends of the
sleeve need to have substantial expansions into the parent tube and the sleeve needs to have
10937790
441
adequate wall thickness to match the heat exchanger conditions. By sleeving tubes, rather than
plugging them, the reliability of the heat exchanger performance can be maintained. This will
result in improved plant performance and will result in less operating stress (in the form of less
flow and heat transfer requirements) on other heat exchangers within the system.
The following sections provide a description of the various types of sleeves that can be used to
structurally repair defective heat exchanger tubes. In all cases, it is crucial that pressure tests, at
design and operating conditions, be performed on qualification mockups simulating the final
field condition. These tests are used to verify that the sleeve design and expansion sizes are
adequate to withstand the heat exchanger conditions.
Tube End Stiffening Sleeve

Tube erosion can occur at the inlet end of softer tube materials, such as brass, copper-nickel, or
carbon steel alloys. As the tubes thin, the residual stress that was once present to hold the tube
against the tubesheet is relieved and leakage can occur across the seal joint. Because of the
location of the thinning, it is very difficult to detect damage by methods other than a shell side
pressure test. In addition, traditional tube plugging may not seal the leak and rerolling the
leaking joint may result in increased leakage due to further tube thinning. While the leak rate at
each location may be small, the overall leakage can be excessive since it can occur in a large
number of locations since many of the tube ends may thin at the same rate. Also, a sudden
pressure or temperature change in the heat exchanger could result in a rapid increase in leakage.
Figure 1 shows some thinned tube ends in a water-to-air cooling coil.
Figure 1
Thinned Tube Ends in Heat Exchanger
A solution to this is to install a stiffening sleeve which can serve two purposes:
1) Provide reinforcement for the existing thinned roll expansion. When the stiffening sleeve
is roll expanded into the tube, it will not only hold itself in place, but will also expand the
parent tube into the tubesheet. This will place the tube expansion back into its original
leak tight configuration.
2) Minimize further inlet end parent tube erosion, eliminating the risk of future tube wall
thinning. The inside of the tube is shielded from flow through the installation of the
sleeve. The sleeve should be fabricated from a material which is harder than the original
tube to assure that future damage does not occur.
442 10937790
Mockup assemblies, such as those shown in Figure 2, should be fabricated from the parent tube
and tubesheet material in order to perform these tests. The tube ends should be thinned to
simulate the existing heat exchanger condition. Afterwards, the tube-to-tubesheet joint is
pressure tested to determine the amount of leakage present prior to sleeve installation.
Figure 2
Qualification Mockup Sample
After sleeve installation at various torque values, to simulate the upper and lower limits, the
tube-to-tubesheet joint is pressure tested to assure that the leakage has been eliminated. The final
acceptable torque is one that stops leakage without applying undo stress to the tubesheet
ligament.
Stiffening sleeves can be utilized within any heat exchanger that has thinned tube ends and that
may be leaking. These include low pressure heat exchangers such as condensers, turbine and
reactor building service water coolers, and air coolers. In addition, feedwater heaters which have
tubing manufactured from soft materials (such as carbon steel and copper-based alloys) can also
be repaired through the use of a stiffening sleeve. While the feedwater heater pressures may be
much higher, a structural sleeve installed into the tube end can provide additional backing to
limit or eliminate further tube-to-tubesheet leakage. The use of a stiffening sleeve can
significantly extend the life of a heat exchanger that is suffering from tube end wear loss and
leakage.
Field Experience
At Oconee Nuclear Station, during the Fall 2004, Unit 3 outage, leakage was identified at the
tube-to-tubesheet joint in the reactor building cooling units (RBCUs). These water-to-air heat
exchangers consist of 12 coils per Unit, with 96 u-tubes per coil. Figure 3 shows one of the
coils. The visual indication of leakage, along with the known thinned 90-10 copper-nickel tube
ends, resulted in a decision by Duke Energy to install tube end stiffening sleeves during
subsequent Unit 1, 2, and 3 outages (time was limited during the Unit 3 outage and sleeving was
deferred until the next outage). Qualification testing, performed prior to the Unit 1 outage,
showed that thinned tubes (with 66% wall loss in the roll expansion area) that had leak rates as
high as 0.006 gallons per hour were completely sealed after sleeve installation. The hydrostatic
leak tests were performed at design, normal operating, and a waterhammer transient pressure.
Axial load tests were also performed to verify that the sleeves would not move if they were
plugged and the full waterhammer pressure was applied to the plug and sleeve.
Framatome ANP installed 374 stiffening sleeves (1/3 of all tubes) into the RBCU inlet tube ends
during the Oconee Unit 1 Spring 2005 outage. Sleeves were installed into tubes with the greatest
end erosion damage which were located in front of the inlet nozzle feeding the waterbox.
10937790
443
Figure 3
Oconee Unit 1 RBCU Coil (One of Twelve)
Tubesheet Sleeve
Many times, tube damage can occur near the inlet or outlet tubesheet. This damage may be in
the form of cracking at the tube roll expansion transition, cracking or tube OD pitting on the shell
side of the tubesheet, flow erosion damage away from the tube end, or tube ID pitting near the
tube end due to impurities from improper layup. It may be possible to bridge any of these flaw
types with a relatively short structural sleeve, allowing the tube to remain in service.
Variations of the tubesheet sleeve exist and the actual type that is used will depend on the heat
exchanger operating conditions and the axial location of the damage. The types of sleeves are:
• Short sleeves that are roll expanded on each end. The tubesheet expansion is torque
controlled and the freespan expansion is diameter controlled. The settings for both
expansions are set so that they can provide both the structural and leakage requirements for
the sleeve.
• Short or long sleeves that are roll expanded on each end and hydraulically expanded over
their full-length. The addition of the full-length hydraulic expansion is used if the loss of
heat transfer is an issue (with a longer sleeve) or if any restriction within the tube ID is
undesirable (if an automatic tube cleaning system is used). While the full-length expansion
is typically not a structural or leak limiting joint, it does improve the heat transfer and flow
through the sleeved tube. The full-length expansion is made by placing a tool, with seals on
each end, into the sleeve. The inside of the sleeve is filled and then pressurized with water,
expanding the sleeve into tight contact with the tube.
• Short or long sleeves that are roll expanded in the tubesheet, hydraulic expanded on the
freespan end, and hydraulically expanded over their full-length. The use of a hydraulic
expansion on the freespan end can improve the leak resistance of the sleeve. A sleeve of this
type would typically be used in a higher pressure heat exchanger, such as a feedwater heater.
The hydraulic freespan end expansion is made with a tool that has multiple plastic bladders.
These bladders, which are filled with high pressure water, force the sleeve into the tube until
the proper size expansion is made (as determined by a control computer).
444 10937790
Qualification tests are used to set the sleeve installation parameters and may include the
tubesheet roll expansion torque, the freespan roll expansion size, the full-length expansion
pressure, or the freespan hydraulic expansion constants. Mockup assemblies must be fabricated
which simulate the tube and tubesheet condition and materials. After installing sleeves using the
upper and lower expansion limits, the sleeves should be pressure tested and the actual leakage
through the end joints measured. The expansion type used at the sleeve ends must match the
allowed leakage for the heat exchanger application.
Field Experience
The use of short length structural tubesheet sleeves can reliably extend the life of heat
exchangers with localized damage near the tube end. Framatome ANP has installed these
sleeves into condensers (Figure 4), feedwater heaters (Figure 5), and component cooling water
heat exchangers (Figure 6). A wide range of parent tube materials has been sleeved, including
AL-6X, Monel, and copper-nickel. The tube side operating pressure has ranged from low, in the
case of condenser sleeves, to high, in the high pressure feedwater heaters at a fossil plant.
Figure 4
Condenser Tubesheet Sleeving (Sleeve Installation and Full-Length Expansion)
Figure 5
Component Cooling Water Heat Exchanger Tubesheet Sleeving (Roll Expansion)
10937790
445
Figure 6
Feedwater Heater Tubesheet Sleeving (Insertion and Roll Expansion)
Deep Bundle Sleeve

This type of sleeve is used to repair defective areas of tubing located deep in the bundle, away
from the tube end. Many times, the damage mechanism is tube wear, due to excessive vibration,
at tube support locations. This sleeve can be installed anywhere along the straight length of the
tube.
The sleeve is expanded into the tube by a hydraulic expander. This expander, shown in Figure 7
along with a representative sleeve and tube, consists of multiple plastic bladders that are filled
with high pressure water. As the water pressure increases, the bladders expand against the inside
of the sleeve, pushing the sleeve into the tube. The expansion process, which is computer
controlled, continues until a preset volume of water is reached. At this point the sleeve is
properly expanded and the bladders are depressurized.
Qualification testing is performed to set the hydraulic expansion parameters. After installing the
sleeves, the leakage through the end joints should be measured, with a requirement that the leak
rates must match the allow leakage for the heat exchanger application.
Field Experience
Deep bundle sleeves have been used to reliably extend the life of heat exchangers with localized
damage deep within the tube bundle since 1993. Framatome ANP has installed these sleeves
into component cooling water (Figure 8) and containment spray (Figure 9) heat exchangers and
feedwater heaters. While the majority have been installed into stainless steel tubes, sleeving has
also been performed in carbon steel and copper-nickel tubing.
Figure 7
Deep Bundle Expander, Sleeve, and Tube
446 10937790
Figure 8
Component Cooling Water Heat Exchanger Deep Bundle Sleeving
Figure 9
Containment Spray Heat Exchanger Deep Bundle Sleeving
Full-Length Sleeve
These sleeves are installed to span the entire length of straight-tubed heat exchangers. After
insertion, the full-length of the sleeve is expanded into the parent tube. This step serves the dual
purpose of maintaining heat transfer as high as possible (typically 75%-90% of the parent tube)
while minimizing flow pressure drop through the tube. After the full-length expansion step, the
sleeve ends are trimmed flush with the existing tube ends and the sleeve is roll expanded into the
tubesheet.
Full-length sleeves are typically used in condenser or cooling water heat exchangers when there
are multiple defects along the tube length. These sleeves are an attractive option when a
relatively small percentage of tubes require repair or tube replacement is difficult as a result of
the location of the heat exchanger within the plant. Through sleeving, the majority of the tube’s
heat transfer area can be left in service, resulting in a heat exchanger that operates at close to its
as-designed condition.
10937790
447
Full-length sleeving is comparable in many ways to retubing in the methods employed to install
the sleeves. However, since removal of the existing tube is not required, and the typical number
of tubes that will be full-length sleeved are below the number that would be retubed, the cost for
material and manhours are much less than for retubing. This makes sleeving a cost-effective
option to return and keep tubes in service.
These sleeves require qualification testing to set the roll expansion torque and full-length sleeve
expansion pressure. Flaws that simulate those found in the heat exchanger should be placed in
the qualification tubes during full-length expansion testing.
Since the sleeve covers the entire parent tube length, this type of sleeve can have the longest
repair life of the four. However, for this type of sleeve to be effective it must have an adequate
wall thickness and the material type must be properly chosen for the existing service water
conditions.
Field Experience
The use of full-length structural sleeves can reliably extend the life of heat exchangers with
defects located at many points along the tube length. As shown in Figure 10, Framatome ANP
has installed these sleeves into condensers with copper-nickel parent tubing.
Figure 10
Condenser Full-Length Sleeving (Insertion and Full-Length Expansion)
ITEMS TO CONSIDER FOR TUBE REPAIR

Prior to performing tube sleeving, the following factors should be considered:
• The length, location, and quantity of tube defects that would require sleeving need to be
determined.
• Whether on- or off-line tube cleaning is performed. In either case, the sleeve needs to be
sized and expanded to assure that the cleaning devices do not become stuck during use.
• The proper sleeve material must be used. The sleeve needs to be compatible with the heat
exchanger parent tubing and the tube and shell side water chemistry. The galvanic corrosion
potential and the effects of crevice corrosion between the sleeve and tube, in the heat
exchanger water chemistry, need to be considered.
448 10937790
• The sleeve needs to be sized for the heat exchanger design and operating conditions,
including any ASME Code minimum wall thickness calculation. The sleeve must be long
enough to span the tube defects, but needs to fit within any tube end clearance restrictions.
• It may be necessary to calculate the heat transfer and flow loss due to installing a large
quantity of sleeves. These calculations will give a sleeve-to-plug ratio that can be used to
determine the effect on heat exchanger performance after sleeving is complete (and tubes
have been returned to service, if applicable).
Based on the results of the study, a decision can be made on whether sleeving could be an
effective repair for the heat exchanger. If sleeving is possible, it can provide a long term solution
to tubing operability issues.
CONCLUSIONS
The costs associated with heat exchanger replacement can be significant. These costs include the
new heat exchanger, tube bundle, or retubing effort, the manpower required to remove the old
and install the new heat exchangers or tubes, plant modifications to allow for the removal of the
heat exchanger or tubes, and the amount of outage time associated with replacement. In
addition, the replacement of a heat exchanger or its tubing can adversely affect other work going
on in the vicinity of the heat exchanger. Because of the cost and time involved, and if the
damage is confined to only the tubing (which is typically the case), repair of the heat exchanger,
through structural sleeving, should be considered. If the tube damage is confined to one general
area, there is a good possibility that the expense of a replacement can be avoided. In addition,
the time required to prepare for tube repair is much less than for replacement (2-8 weeks
compared with 18 months), allowing a decision on repair to be made just before, or even into, an
outage.
By removing plugs and installing sleeves, it is possible to return lost heat transfer area to service.
Tubes that would be likely to fail in the near-term can also be repaired. This will improve the
performance and reliability of the heat exchanger. The cost to perform the repairs is also much
less than for replacement (usually less than 1/10th the cost). Sleeving has been shown to be a
proven tube repair technique, having been performed since the 1970s, with some sleeves in
service for more than 15 years. During this time, tube repairs have economically extended the
useful life of heat exchangers worldwide.
As the number of plugged tubes approaches unacceptable levels, or if damage is consistently
occurring in one area of a heat exchanger, tube sleeving should be considered to minimize future
damage and extend the life of the heat exchanger.
REFERENCES
1. NMAC Feedwater Heater Maintenance Guide. Charlotte, NC: Electric Power Research
Institute, May 2002. EPRI 1003470.
10937790
449
10937790
Long Term Restoration to Address Condenser Tube Inlet End
Erosion/Corrosion
EPRI
August 31-September 1, 2005
The Catamaran Hotel

Prepared by: W. H. Stroman

Primary Energy Company
Formally with: Aptech Engineering Services, Inc., Sunnyvale, CA 94089
DISCLAMER
This report was prepared by William H. Stroman, while employed with Aptech Engineering Services, Inc.
and (hereafter called the Company) for oral presentation to EPRI-Condenser Technology Conference.
Neither EPRI nor the Company, or others involved in the preparation or review of this report, nor any of
their respective employees, members or persons acting on their behalf, makes any warranty, expressed or
implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any
information, apparatus, product, or process disclosed, or represents that its use would not infringe privately
owned rights.
Reference herein to any specific commercial product, process, or service by trade name, trademark,
manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendations, or
favoring by EPRI, the Companies, or others involved in the preparation or review of this report, or agency
thereof. The views and opinions of the authors, contributors, and reviewers of the report expressed herein
do not necessarily reflect those of EPRI, the Company, or others involved in the preparation or review of
this report, or any agency thereof.
10937790
451
Abstract:
During 1982, a West coast utility unit was experiencing frequent condenser tube leaks in
the inlet end of copper alloy tubes in a seawater-cooled condenser. These leaks were
identified to be the result of inlet end erosion/corrosion. Based on measurements of a
statistically significant number of tube samples, a high numbers of tube failures were
projected to occur in the near future. Three alternatives were evaluated to remedy the
problem:
1. Retube the condenser
2. Install inlet end protective shields
3. Epoxy coat the inlet end of the tubes
All tube samples pulled prior to 1982 had shown little or no erosion/corrosion in areas
other than the inlet end, even after more than 14 years service. The same pattern was also
noted in other units at the station with up to 20 years service. Since the causes of the
degradation were isolated to a specific section of the condenser, it was expected that the
inlet shields would be effective in getting a proper seal between the tube and tube sheet
and eliminate the inlet end erosion problem. It was also believed that, with properly
maintained cathodic protection and ID tube cleanliness program, the shields would not
cause any harmful side effects to the tubes or tube sheet.
A tube specimen was removed during the 2004 maintenance outage to determine the
failure mechanism of a leaky tube that had been plugged prior to the outage. The leak
was determined to be the result of through-wall damage from steam impingement that
had occurred near the exit end of the condenser tube. Inspection of the tube section that
contained the sleeve insert confirmed that the inserts were performing as intended by
providing inlet end protection of the condenser tubes, allowing for long-term restoration
of the condenser tubes.
Background:
The seawater-cooled condenser consists of 1-inch OD, 20 BWG 90-10 CuNi tubes and
Muntz metal tube sheet with zinc waste plates for cathodic protection. Tube failures were
occurring at a frequency of once a week, with multiple tube leaks occurring with
increasing frequency and expected to increase based on trending. In order to prevent
boiler damage, leak detection expenses and lowered unit availability, the installation
needed to be done in the most expeditious manner possible. Highlights of the three
alternatives evaluated are as follows:
Retubing:
• Most reliable method
• Most expensive – in terms of installed cost, costs were 6.41 times for sleeve
inserts and 8.33 times for epoxy coating.
• Most outage time – 4weeks unit had to be off line.
452 10937790
Epoxy:
• Least long term reliability due to brittleness of coating and potential for pin hole
leaks
• Concern that the use of mechanical scrapers to periodically clean the condenser
tube ID surfaces of deposition and marine growth may damage the epoxy coating.
Maintaining the cleanliness of the waterside of the condenser tubes is essential in
prolonging the life of the tubes as well as in maintaining condenser performance.
• Least expensive in terms of installed cost and outage cost, ~30% less than
installing sleeve inserts
• Least outage time – 1 week unit had to be off line.
Inlet end Shields (Metal Inserts)

• Industry experience had noted improper or lack of cathodic protection (a must
with seawater cooling) could result in localized pitting at the downstream
transition where the shield ends ~6” into the condenser tube. An example of
failure associated to galvanic corrosion is shown in Figures 1 and 2.
• Only slightly more expensive than epoxy coating
• Could accommodate metal and brush scrapers used to periodically clean the
condenser tube ID surfaces of deposition and marine growth. Maintaining
condenser cleanliness is also a requirement for maintaining condenser
performance.
• Least outage time – 1 week. With condenser half out of service could be done on
line.
Inlet end Shields (Plastic Inserts)

Plastic inlet inserts were another option but industry experience was not favorable for
long-term restoration and was not considered as:
• They tended to provide protection for a period of 2-10 years
• The thickness at the inlet the sleeve results in an ID reduction of around 35% at
the tube entry
• The plastic inserts are glued in versus expanded into place, as is the case with the
metal sleeve. Plastic inserts do not insure correction of the inlet end corrosion
parent tube-to-tube sheet integrity, as is the case with metal inserts.
• Plastic liners tend to loosen and work free of the tube over time. The void around
the lip of the inlet end of the insert has resulted in severe tubesheet ligament
damage to tube sheets.
• The 35% entrance reduction inhibits cleaning the tubes with any properly sized
cleaning device, as nothing that goes through the insert has a high likelihood of
cleaning and maintaining clean condenser tube ID surfaces. Maintaining the
cleanliness of the waterside of the condenser tubes is essential in prolonging the
life of the tubes. Poor tube cleanliness will also affect the unit’s overall efficiency.
• If a tube leak develops, it is difficult to get proper seal of the plastic liner to the
tube sheet to address any leaks at this interface.
10937790
453
• Based on problems with plastic inlet sleeves used within the U.S. Navy’s main
ship condensers, all of the plastic inserts were in the process of being removed
from the ship’s condensers and replaced with metal shields.
Short Term versus Long Term Benefits:

All tube samples pulled prior to restoration measures have shown little or no
erosion/corrosion in areas other than the inlet end after more than 14 years service.
Therefore, with an ongoing ID tube maintenance-cleaning program, the minimum
expected service life of the tubes following shield installation should approach or exceed
another 14 years. Due to the relatively low cost of the project and the resultant payback
of less than 9 months and the higher degree of confidence of continued service life of the
tubes versus epoxy coating, the recommendation was to have CTI Industries install
AL6X shields.
Twenty-Two Years Later:

During 2003, the condenser had developed a condenser leak. It was the first leak to occur
in a number of years. Early in 2004 another condenser leak in the same area was
identified and plugged. As part of a scheduled maintenance, the tube that was recently
plugged was pulled to determine root cause failure mechanism. The tube had failed as a
result of OD through-wall pitting from steam impingement near the exit end of the
condenser.
As part of the metallurgy assessment of the removed tube, the tube inlet and protective
shield were also examined. The concern was that during 2000, as part of an
environmental upgrade, the cathodic protection had been changed from zinc waste plates
to impressed current. Of specific concern was the interface between the shield exit end
and the condenser tube where the potential for galvanic corrosion is highest (an example
of sleeve exit end galvanic corrosion is shown in Figures 1 and 2).
There was no evidence of any erosion or corrosion at the exit end of the shield. Figures 3
and 4 show the as-found condition of the inlet tube surface below the shield, as well as at
the shield exit end after cleaning to remove surface deposits and oxides.
Conclusion:
Based on the 22 years of continued service and the findings of the metallurgical
examination, the decision to install protective inlet shields to provide long-term
restoration of the condenser was the correct choice. Effort in maintaining proper cathodic
protection and a maintenance-cleaning program for the tube ID surfaces are essential
requirements for prolonging the life of tubes. The shield’s internal diameter did not
prevent the continued use of mechanical cleaning devices for periodic maintenance
cleaning, ensuring continued condenser performance within design specifications.
454 10937790
Figure 1: An example of galvanic corrosion occurring at the exit end of the shield and the
condenser tube interface due to poor cathodic protection.
Figure 2: Example of a pit that occurred as a result of galvanic corrosion at the exit end of
the shield and the condenser tube interface.
10937790
455
Figure 3: Condenser tube inlet where the AL6X shield had been installed (Note: The
lodged wooden tube plug could not be removed when the tube was pulled).
Figure 4: The tube inlet surface with shield removed and oxide downstream of the shield
removed by abrasive blast cleaning, showing no evidence of exit end corrosion after 22
years service.
456 10937790
10937790
Export Control Restrictions The Electric Power Research Institute (EPRI)
Access to and use of EPRI Intellectual Property is granted with The Electric Power Research Institute (EPRI), with major locations in
the specific understanding and requirement that responsibility Palo Alto, California, and Charlotte, North Carolina, was established
for ensuring full compliance with all applicable U.S. and in 1973 as an independent, nonprofit center for public interest
foreign export laws and regulations is being undertaken by energy and environmental research. EPRI brings together members,
you and your company. This includes an obligation to ensure participants, the Institute’s scientists and engineers, and other leading
that any individual receiving access hereunder who is not a experts to work collaboratively on solutions to the challenges of electric
U.S. citizen or permanent U.S. resident is permitted access power. These solutions span nearly every area of electricity generation,
under applicable U.S. and foreign export laws and delivery, and use, including health, safety, and environment. EPRI’s
regulations. In the event you are uncertain whether you or members represent over 90% of the electricity generated in the
your company may lawfully obtain access to this EPRI United States. International participation represents nearly 15% of
Intellectual Property, you acknowledge that it is your EPRI’s total research, development, and demonstration program.
obligation to consult with your company’s legal counsel to
determine whether this access is lawful. Although EPRI may Together...Shaping the Future of Electricity
make available on a case-by-case basis an informal
assessment of the applicable U.S. export classification for
specific EPRI Intellectual Property, you and your company
acknowledge that this assessment is solely for informational
purposes and not for reliance purposes. You and your
company acknowledge that it is still the obligation of you and
your company to make your own assessment of the applicable
U.S. export classification and ensure compliance accordingly.
You and your company understand and acknowledge your
obligations to make a prompt report to EPRI and the
appropriate authorities regarding any access to or use of EPRI
Intellectual Property hereunder that may be in violation of
applicable U.S. or foreign export laws or regulations.
Program:
Combustion Performance and NOX Control
© 2005 Electric Power Research Institute (EPRI), Inc. All rights reserved.
Electric Power Research Institute and EPRI are registered service marks of
the Electric Power Research Institute, Inc.
Printed on recycled paper in the United States of America 1010322
ELECTRIC POWER RESEARCH INSTITUTE

3420 Hillview Avenue, Palo Alto, California 94304-1395 • PO Box 10412, Palo Alto, California 94303-0813 USA
800.313.3774
10937790 • 650.855.2121 • askepri@epri.com • www.epri.com

Tccnologías Condensador PDF

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Tccnologías Condensador PDF

Transféré par

Droits d'auteur :

Formats disponibles

Condenser Technology Conference

Proceedings, August 2005

EPRI Project Manager

ELECTRIC POWER RESEARCH INSTITUTE

(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I)

(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER

ORGANIZATION(S) THAT PREPARED THIS DOCUMENT

Electric Power Research Institute (EPRI)

This report was prepared by

Electric Power Research Institute (EPRI)

This report describes research sponsored by the EPRI.

Condenser Technology Conference. EPRI, Palo Alto, CA: 2005. 1010322.

CONDENSER TECHNOLOGY CONFERENCE

AUGUST 31-SEPTEMBER 1, 2005

WEDNESDAY, AUGUST 31, 2005

Session 1 Plenary Session I

9:00am Improving The Reliability of A BWR Through A Condenser

9:30am Modification of Extraction Steam Expansion Joints in Low Pressure Condensers

10:00am Break and Exhibits

11:30am The Cost of Dirty Condenser Tubes

Noon Lunch and Exhibits

Session 3 Maintenance Technology I

1:30pm Remote-Field (RFT) Inspection Techniques to Characterize and Size OD

2:30pm Improving Condenser Efficiency With Innovative Scale Removal System

3:00pm Break and Exhibits

5:00pm Hosted Reception and Exhibits

THURSDAY, SEPTEMBER 1, 2005

Session 4 Fouling Control

9:00am The Practical Application And Innovation of Cleaning Technology for

9:30am Technology for Cleaning The External Surface of Air-Cooled Condensers

11:00am AmerenUE Labadie Unit 2 Copper-Beryllium Tube Development of a Cobalt-

Session 6 Maintenance Technology II

1:25pm New Developments In Condenser Tube Coatings

2:15pm Heat Exchanger Tube Structural Repair Sleeves

3:30pm Panel Discussion: Innovations in Maintenance Technology

WEDNESDAY, AUGUST 31, 2005

Condenser Tube Failures – Theory and Practices: An EPRI Manual

Session 2 Performance Technology ................................................................................. 71

On A Predictive Condenser Performance Model

Session 3 Maintenance Technology I.............................................................................. 107

Eddy Current Application for Balance-of-Plant Heat Exchangers

THURSDAY, SEPTEMBER 1, 2005

Session 4 Fouling Control................................................................................................ 229

Condenser Biofouling at Entergy’s Gerald Andrus Power Plant Diagnosis and

Session 5 Plenary Session II............................................................................................ 295

Vibration Prevention Goes Beyond Support Plate Spacing

Session 6 Maintenance Technology II............................................................................. 399

Condenser Tube Plugs

Condenser Tube Failures - Theory and Practice:

Kevin J. Shields, EPRI

There is an overwhelming body of information pertaining to damage in condenser

• Serving as an impervious barrier between steam/condensate and circulating

• To provide a single reference of information about condenser tube failure

• To provide direct, specific actions to be taken if a condenser tube failure has

• To provide sufficient background information so that the user can understand

Impact of Condenser Tube Damage/Failures

Effects of condenser tube leaks and damage on steam generators. Steam

Effects of condenser tube damage on plant discharge. Excessive waterside

As a result of these significant plant impacts, it is important that condenser tube

Key Perspectives for Approaching CTFs

• It is paramount to understand and apply the concept that a clear

• Many corrosion modes in the condenser can be prevented or minimized

the water side, damage mechanisms such as sulfide attack, dealloying,