Condenser Technology

Condenser Technology
Seminar and Conference

Technical Report
EPRI Project Manager
J. Stallings
EPRI 3412 Hillview Avenue, Palo Alto, California 94304 PO Box 10412, Palo Alto, California 94303 USA
800.313.3774 650.855.2121 askepri@epri.com www.epri.com
Condenser Technology
Seminar and Conference
1004116
Proceedings, September 2002

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
John Tsou Consulting Services

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. EPRI. ELECTRIFY THE WORLD is a service mark of the Electric Power
Research Institute, Inc.
Copyright 2002 Electric Power Research Institute, Inc. All rights reserved.

iii
CITATIONS
This report was prepared by
John Tsou Consulting Services
56 Williams Lane
Foster City, CA 94404
Principal Investigator
J. Tsou
This report describes research sponsored by EPRI.
The report is a corporate document that should be cited in the literature in the following manner:
Condenser Technology: Seminar and Conference, EPRI, Palo Alto, CA: 2002. 1004116.

v
REPORT SUMMARY

Steam surface condensers and associated systems cause significant loss of availability and heat
rate degradation in both nuclear and fossil-fired power plants. Nineteen papers, presented at a
2002 Conference, discussed industrial experience and case histories of condenser problems and
solutions.
Background
Power plant efficiency and availability depend greatly on condenser performance. However,
improvements in operation and maintenance practices, materials, and design may correct most of
these losses. To address these problems and identify solutions, EPRI has brought together utility
representatives, consultants, and manufacturers in seven conferences on condenser technology
since 1981. Subsequent to the last EPRI conference on condenser technology, which was held in
September 1999 (EPRI report TR-106781), EPRI, utility members, equipment manufacturers,
and others have done significant work to improve system reliability, performance, and life
extension of this equipment and associated systems. The current report on CD is the proceedings
of the eighth EPRI-sponsored conference on condenser technology.
Objectives
To review and document condenser and associated system problems, solutions, and
improvements.
Approach
Participants, including utility engineers, manufactures, and consultants attended a conference on
condenser issues from September 10-11, 2002, in San Antonio, Texas. During six formal
sessions, 18 technical presentations covered condenser problems and solutions. Two panel
discussion sessions provided a forum for exchanging information in a less formal environment.
At the exhibition, vendors displayed the latest materials, equipment, and services. Participants
could also attend an optional seminar held before the conference.
Key Points
This CD contains papers on condenser design, material, maintenance, performance, fouling
control, and alternative cooling technology. To assist utility personnel in reducing operation and
maintenance cost, the report documents the latest experience on condenser problems and
solutions. Presentations covered the following topics:

Improvements in condenser design and upgrade
Case histories of corrosion resistant tube materials

vi
The impact of air in-leakage and fouling on performance
Case histories of fouling control technology
Alternative cooling technology
A condenser performance testing procedure
Keywords
Condensers
Maintenance
Fouling
Corrosion
Failure Analysis
Performance

vii
ABSTRACT

A conference on condenser issues was held from September 10-11, 2002, in San Antonio, Texas.
Participants included utility engineers, manufactures, and consultants. During six formal
sessions, 18 technical presentations covered condenser problems and solutions.
Three papers discussed the improvements in condenser design and upgrade.
Two papers documented case histories of corrosion resistant tube materials.
Four papers documented condenser maintenance technologies.
Four papers discussed the impact on performance of air in-leakage and fouling.
Three papers described case histories in fouling control technology.
Two papers were related to alternative cooling technology.
One paper documented a condenser performance testing procedure.
Two panel discussion sessions provided a forum for exchanging information in a less formal
environment. At the exhibition, vendors displayed the latest materials, equipment, and services.

ix
CONTENTS
SESSION 1: DESIGN TECHNOLOGY ................................................................................... 1-1
Power Uprate: The Engineering Evaluation of the Condenser and Cooling Tower ............. 1-3
D. C. Burns, R. L. Stevens, J. M. Burns, Burns Engineering Services Inc.
CFD Analysis Predicts Condenser Performance after Large Power Uprate of the
Quad-Cities and Dresden Stations ....................................................................................1-17
N. Rhodes, C. D. Hardy, Heat Exchanger Systems Inc.
J. M. Burns, Burns Engineering Services Inc.
T. B. Madden, Stone & Webster Inc.
Surface Condensers, Steam Dump, & Design Reliability Considerations ..........................1-29
David H. Cooley, ALSTOM Power
SESSION 2: MATERIALS TECHNOLOGY............................................................................ 2-1
Thin-Wall Titanium Condenser Tubing: The Next Plateau .................................................. 2-3
Dennis J. Schumerth, Valtimet, Inc.
The Performance of Superferritic Stainless Steels in High Chloride Waters ......................2-17
John C. Tverberg, Edward R. Blessman, Trent Tube
A Main Condenser Success Story .....................................................................................2-29
Jim Mitchell, Plastocor, Inc.
SESSION 3: MAINTENANCE TECHNOLOGY....................................................................... 3-1
Six Case Histories of High Reliability Coatings and Condenser Corrosion Problems.......... 3-3
Michael J. Horn, James E. Mitchell, Plastocor, Inc.
The Use of 100% Solids Epoxy Coatings for Full Length Condenser Tube Linings ...........3-15
A. Wesley Langeland, Duromar, Inc.
Richard Kreiselmaier, Plastocor-international SA
Bruce Woodruff, Progress Energy
Heat Exchanger Tube Side Maintenance Repair vs. Replacement.................................3-19
Bruce W Schafer, Framatome ANP, Inc.
Condenser Application and Maintenance Guide................................................................3-35
Sharon R. Parker, Contractor
Alan Grunsky, EPRI

x
SESSION 4: PERFORMANCE IMPROVEMENT.................................................................... 4-1
A Condenser Success Story............................................................................................... 4-3
Dave Leissner, Mirant
Richard Putman, Conco Consulting Corp.
On Understanding Condenser Pressure Saturation at Low Air Ingress..............................4-17
Joseph W. Harpster, Intek, Inc.
Experiences With Steam Cycle Air Ingress at Alliant Energy Fossil Plants: Case
Studies Documenting the Effects of Air In-Leakage on Steam Plant Operations................4-27
Wesley A. Kaufman, P.E., Alliant Energy IP&L
The Measurement of Condenser Losses Due to Fouling and Those Due to Air Ingress ....4-43
Richard E. Putman, Conco Consulting Corp.
Joseph W. Harpster, Intek Inc
SESSION 5: FOULING CONTROL ........................................................................................ 5-1
SIDTEC Condenser Cleaning for Cooling Water Systems.................................................. 5-3
R. Jones, S. D. Jones, R. Post, GE Betz
J. F. Echols, SIDTEC Services, Inc.
On-Line Automatic Tube Cleaning System and On-Line Self Flushing Debris Filter ..........5-19
Kaveh Someah, Brackett Green WSA, Inc
Mechanical Tube Cleaning: A Brief Tutorial .......................................................................5-29
George Saxon, Jr., Conco Systems, Inc.
SESSION 6: ALTERNATIVE COOLING TECHNOLOGY....................................................... 6-1
Wet And Dry Cooling---Cost/Performance Tradeoffs .......................................................... 6-3
John S. Maulbetsch, Maulbetsch Consulting

Kent D. Zammit, EPRI
Matthew Layton and Joseph OHagan, California Energy Commission
The Impact of Air Cooled Condensers on Plant Design and Operations............................6-19
Dirk Jaresch, J & W GmbH
Experience Evaluating Condenser Performance and Tube Fouling with the ASME
Performance Test Code on Steam Surface Condensers PTC 12.2 -1998 .........................6-35
Dale C. Karg, Santee Cooper
John M. Burns, Burns Engineering Services Inc.
Michael C. Catapano, Powerfect Inc.

1-1
1
SESSION 1: DESIGN TECHNOLOGY
Power Uprate: The Engineering Evaluation of the Condenser and Cooling Tower
Daniel C. Burns, Robert L. Stevens, John M. Burns
Burns Engineering Services Inc.
CFD Analysis Predicts Condenser Performance After Large Power Uprate of the Quad
Cities and Dresden Stations
N. Rhodes, C. D. Hardy, Heat Exchanger Systems Inc.
J. M. Burns, Burns Engineering Services Inc.
T. B. Madden, Stone & Webster Inc.
Surface Condensers, Steam Dump, & Design Reliability Considerations
David H. Cooley
ALSTOM Power

Session 1: Design Technology
1-3
POWER UPRATE:
THE ENGINEERING EVALUATION OF THE
CONDENSER AND COOLING TOWER
Daniel C. Burns
Robert L. Stevens
John M. Burns
PO Box 272
Topsfield, MA 01983-0272

Abstract
Today many nuclear and fossil based utilities have either completed or intend to appreciably
increase the generation of their existing power plants. Power uprate and other associated terms
such as repowering, modernization, refurbishment or upgrade, are often used to describe this
type of project. Modernization of existing nuclear & coal facilities has been found to be a cost
effective way to maximize asset value, especially when compared to installing new generation.
The power uprate strategically supports the deregulated market based competitive initiative of
increasing production value while lowering costs.
The power cycle equipment often has an extra margin or can be redesigned to allow operation at
the higher power level. However, the power uprate produces an increase in the heat duty to the
existing condenser and any closed-cycle cooling tower. Uprates have produced as much as a
35% increase in the turbine exhaust steam flow above the original design basis of the condenser.
Often uprate projects have been executed with only minimal consideration of the essential
detailed engineering of the impacts on the cooling system and related components. Important
effects of the uprate on the condenser include excessive turbine backpressures during peak
operation, high condensate temperatures, as well as exhaust pressure levels that preclude
attaining the station generation that was the economic justification for going forward with the
uprate. Other significant cooling system performance effects after an uprate include the
capability of the steam jet air ejector or vacuum pump to remove the extra load of non-
condensable gases. There will also be an increased potential for condenser tube vibration
failures.
In addition, after a power uprate the cooling towers will experience higher approaches to the
ambient wet bulb temperatures and warmer water returns to the condenser, hotter station
blowdown to the environment, and an increase in evaporation that results in higher make-up
flows. Within the paper, an overview of all these far reaching effects of the power uprate on the
existing condenser and cooling tower will be presented. Simplified but accurate methods for
estimating their increased impacts will be described. The paper will touch on the use of modular
bundle replacement condensers and cooling tower augmentation to reduce the uprate impacts in
extreme applications.

1-4
The paper will provide utilities with a broad variety of quantitative cooling system evaluations
and contribute to the realistic assessment of the financial benefits of adopting large power
uprates for their existing nuclear or fossil plants.
Introduction
In todays climate of deregulation, there are attractive financial incentives for power uprates.
Power uprates increase the plant generation by an upgrade of existing equipment. Generally, the
mechanical equipment in the power cycle has an inherent added capacity for reliable operation at
increased power levels or may require modifications to accommodate these added requirements.
As a result, the capital & opportunity costs associated with a power uprate may be minimal,
while the extra revenues of the station can be appreciable, particularly during peak demand
periods.
Taking advantage of the power uprate strategy, many nuclear and fossil based utilities have
either completed or plan within the immediate future to increase the generation of many of their
existing power plants. But it is not fully appreciated that waste heat is a natural consequence of
the thermodynamic cycle and so all power uprates produce an increase in the heat duty to the
existing condenser and cooling tower. Indeed, the Second Law of Thermodynamics suggests the
typical Rankine-Regenerative-Reheat cycle, that is the technical basis of many steam plants,
produces almost two units of heat rejected for every added unit of work. As an indication of the
increase in power versus total cooling system heat rejection, a typical fossil or nuclear power
plant operates in a thermal efficiency range of 32% to 42% [5]. The thermal efficiency for
nuclear plants is at the lower end of this range. This natural consequence of the uprate is
sometimes overlooked in the feasibility study of the cooling system compatibility. Uprate plans
can be initiated with only minimal consideration of the impacts of the extra heat and steam flow
on the cooling system components; particularly on the condenser and cooling tower.
Important effects of the uprate on the condenser include excessive turbine backpressures during
peak operation, high condensate temperatures, exhaust pressure levels that preclude attaining the
generation which was the justification of the uprate business case, performance effects on the
steam jet air ejector or vacuum pump non-condensable removal equipment. Not to mention the
increased potential for outages caused by tube impingement & vibration failures, especially
during the winter. Further, after a power uprate, any stations with cooling towers will experience
hotter blowdown to the environment and more evaporation that will require higher make-up
flows. Incrementally higher approaches to wet bulb temperatures will also cause warmer water
returns to the condenser and additionally raise the condenser pressure.
For power uprates that seek a large increase in generation, this paper will indicate why simple
estimates to quantify the effects of the extra steam on the condenser performance often are not
adequate. This paper will outline the quantitative estimates of the uprate performance that can be
made of the condenser and cooling towers to accurately determine if revision to the uprate
generational expectations will be required. This paper will also discuss condenser and cooling
tower performance enhancements that can improve the accommodation of a plant to a large
power uprate. This paper does not address the many other effects on the station that must be
considered during a power uprate that are not directly related to the main cooling system.

1-5
Evaluating Existing Cooling System Thermal Performance
Power uprates are applied to older fossil and nuclear plants. Normally, peak summertime
operation is the major basis for the associated business case evaluation. Though there are
exceptions, the condenser usually has some tubes plugged and is also not generally performing at
its design level of apparent cleanliness. Similarly, the cooling tower is often not operating at a
100% design capability and the circulating water pump capacity will likely have slightly
degraded over the years. These effects independently reduce the cost effectiveness of the planned
power uprate on the expected generation.
All installed cooling systems have a finite capability. Unless the anticipated power increase is
low enough so that its impact on the cooling system is very small, the first step is to
quantitatively baseline the current performance levels of the condenser and cooling tower. This
presumes that the condenser pressures at the summer, peak load condition are currently well
below the turbine backpressure limits. An example of a smaller uprate that would likely have a
negligible effect might be one under 2% of the present generating capacity. Many plants have
accurate cooling system instrumentation to monitor cooling system performance parameters in
order to maximize generation and obtain the best heat rate for every specific climate and cooling
water inlet condition [1]. Good records of existing performance must be employed as the
engineering basis for examining the systems ability to accept more heat and steam flow. The
impact of the uprate on the system performance can then be quantitatively evaluated to determine
potential pinch points that may cause significant reductions in gross generation from the
expected uprate levels. Subsequently, any need to consider system modifications can be
definitively addressed.
Baselining the existing unit performance from the design basis is critical. The main parameters
are cooling water flow, the apparent cleanliness of the condenser, and a recent capability of the
cooling tower along with the incidence of entering wet bulbs and inlet water temperatures to the
condenser. If accurate performance information on that equipment is not available, it is
recommended it be obtained. The data is required for any accurate uprate projection going
forward.
Predictions of the potential level of uprate achievable during peak demand periods should be
assessed. The cooling system performance can be determined from a review of historical station
generation derate incidences, available shift data and reasonably accurate operation or test data
developed from the elements of codes such as ASME PTC 12.2 [2] and ASME PTC 23 [3].
Tests of the circulating water (CW) pumps or other CW flow measurements are also required to
determine existing pump capability.
The existing turbine backpressure limits (or the expected limit should it be redesigned), the
incremental turbine characteristic curve, the opportunity cost of the added revenues and the
annual statistical hours of incidence are all needed to fully evaluate the annual extra power, heat
rate, revenues and potentials for unit derating. With the increased heat and steam flow from the
planned power uprate, the baseline performance of the condenser & cooling tower and the
maximum weather/inlet water conditions expected at the site, an estimate of the performance
signature of the cooling system after the uprate can commence.

1-6
The Heat Exchange Institute Condenser Algorithm
Condensers are designed to achieve a certain performance by determining the surface area, and
number of tubes from a variety of conditions such as the quantity of cooling water, steam flow,
space and tubing parameters. These then define the performance estimated by an industry
algorithm from the Heat Exchange Institute [4]. The latter is a tool used to determine the basic
design requirements. The HEI estimate is reasonably accurate and importantly puts all
manufacturers on the same technical basis during the design phase of a project. In effect, it levels
the playing field for the Owners allowing them a simple gauge to measure performance. After
the numbers of tubes are established, the condenser designer develops the detailed tube bundle
pattern, layer by layer based on the space available under the turbine and a host of design rules
and requirements. The manufacturer does not use HEI directly for the many estimates, but
instead relies on empirical test data & correlations that have been developed over the years.
Avoiding excessive steam side velocities, excessive condensate inundation and steam pressure
losses from packing tubes too closely together are paramount design considerations. The
designer will use steam lanes and spacing in a distinct pattern to mitigate higher velocities.
Because the steam is saturated, pressure drops in a condenser tube bundle are always
accompanied by a loss in steam saturation temperature. If the pressure drop is more than
anticipated by the design rules, the temperature loss reduces the effective potential difference
between the steam and the cooling water temperatures. Since all of the steam is condensed, any
extra temperature reduction within the tube bundle causes a backpressure rise that exactly
compensates for the loss. This is a direct penalty to the condenser performance and the target
design backpressure will not be achieved because in effect the overall tube spacing in the bundle
is tight.
In most situations, the HEI method provides a reasonably accurate estimate of the performance
of a specific condenser design during its operation. This method supplies a simple, quick
empirical method of evaluation and has been successfully used since 1932. Though the basic
method has not changed much since then, the correction factors and the heat transfer data has
continually been refined throughout the years to help improve its correlation accuracy.
But one underlying engineering premise must be fully understood when the HEI method is
applied to determine the performance. As inferred in the paragraph above, the calculation of the
HEI heat transfer coefficient fundamentally assumes the tube bundle reflects a pattern of tubes
that were properly located for that design condition. If for instance, the tubes were clustered so
tightly that steam would not be able to easily enter the bundle, the actual performance would fall
very short of the HEI prediction. As a second example, if a tube pattern were constricted so that
the steam did not flow uniformly over all tubes as it moved toward the air cooler section, the
estimated HEI performance would not be realized. And if the tubes of a certain tube bundle were
positioned too close together such that the steam velocities became very high producing higher
than expected steam pressure losses and perhaps significant accumulated condensate rain over
the tubes, the HEI heat transfer coefficient again would not be achieved.
Similarly, the HEI method cannot be applied to variations in the steamside conditions that are
characteristically different than the design conditions because the HEI calculation focuses only
on changes in the waterside heat transfer conditions. Since the waterside resistances generally

1-7
dominate the overall heat transfer, this is a reasonable engineering assumption that allows
simplifications and at the same time provides accurate predictions of the performance of a given
design. Unfortunately though, the HEI method does not accurately forecast the performance of
the condenser when steam flows and velocities increase appreciably.
Thus, it should be recognized that the existing condenser design that is being used in the power
uprate has a tube bundle pattern that was tailored for a maximum quantity of steam. If the uprate
steam flow is much greater than the design basis, HEI will produce a very optimistic prediction
of the condenser backpressure.
Alternative Evaluation of Condenser Performance
To estimate the condenser performance at appreciable power uprate levels, an adaptation of the
well-known thermal resistance technique can be used. The resistance method is more
scientifically based than the HEI algorithm and includes a separate value for the steamside
resistance at the original, design condition. The condenser steamside resistance change that
reflects the added steam flow of the uprate can be estimated from the Nusselt equation with
corrections for bundle steam pressure losses and condensate inundation.
The evaluation procedure for power uprates with their increases in steam flow into an existing
condenser design is an adaptation from Section 5 of the ASME Performance Test Code [2]. The
technique extends the ASME procedures to reflect condensate inundation and incorporates
additional factors for the natural steam-side, two-phase pressure loss effects. The overall
coefficient of heat transfer is estimated from the individual resistances and then used along with
the CW flow, temperature rise, shell surface area and average inlet water temperature to establish
the final condenser pressure. This evaluation procedure has continually evolved from a simplistic
model of the resistance method to more complex models that incorporate condensate inundation
& steam pressure losses.
The major aspects of the detailed computational basis can be found in ASME PTC 12.2, which is
also summarized in a 1994 condenser paper [1]. As an outline, the original design point of the
condenser includes the cleanliness and with that information and the overall heat transfer
coefficient of that design, the fouling resistance (in Units Hr-Ft
2
-F/BTU) can be quantified. By
more precisely determining the waterside resistance from the Rabas-Cane correlation, the
designers original, overall shell side resistance can then be extracted from the computation. The
shell side resistance is then incrementally adjusted for the increased steam flow associated with
the power uprate, condensate inundation and the increased steam pressure losses. Summing and
inverting all the additional resistances associated with the power uprate, produces the overall
condenser uprate heat transfer coefficient. This modified heat transfer coefficient is utilized to
determine the increased condenser pressure during the peak uprate conditions or other situations
of interest.
As had been indicated earlier, compared to the original design basis, the extra steam associated
with a power uprate can be significant. The turbine exhaust pressure upon which any operating
pressure limits and generation are based is exactly at the turbine exhaust, not at the condenser
tube bundle. Steam flows at high velocities from the annulus of down exhaust turbines, turn and
flow over numerous bracing, heaters, extraction lines, the general structure of the exhaust steam

1-8
space and the upper section of the condenser. The disturbances in flow pattern cause a pressure
loss that is higher than the design and this must be also included in the evaluation of the
performance at power uprate. Since the most important point is during peak conditions and this is
usually associated with summer, high absolute pressures, the Mach number of the flow is in the
incompressible range.
Therefore, incompressible formula can be used to estimate this loss. With the computed
condenser tube bundle pressure at uprate conditions, the expected turbine exhaust pressure can
now be estimated for the uprate, compared to any operational limits and from the incremental
turbine characteristic, the generation benefits including added revenues.
An example of the kind of condenser power uprate data that results from an analysis follows in
Table 1. The objective of the study of this cooling system with a multipressure condenser and
natural draft tower was to determine the level of power uprate appropriate, identify any system
pinch points and the expected generation. In this instance, the inlet temperature to the condenser
was first determined from a separate analysis of the cold water approach temperature to the wet
bulb from the natural draft tower using the tower manufacturers curves and recent non-
dimensional capability performance characteristic test data. Notice, in the table below, the high
pressure (HP) compartment turbine exhaust pressures incurred substantial incidences of derate
reducing the benefits of this power uprate to a marginal level.

Table 1
Typical Summary of EPU Effects on Backpressure & Condensate Temperature
Power
Level
Wet
Bulb
Inlet
Water
Turbine
Exhaust
Total
Heat
Turbine Exhaust

Pressure

Condensate
% Uprate
Temp Temp Steam

Load LPZone IP Zone HP Zone Temperature
(Above
existing)
(F) (F) (lbm/hr) (Btu/hr) (in hga) (in hga) (in hga) (F)
0 78 93 8.08E+06 7.72E+09 3.28 4.07 5.36 134.5
5.7 78 93.4 8.54E+06 8.11E+09 3.47 4.32 5.75 138.7
9.5 78 93.7 8.85E+06 8.42E+09 3.61 4.53 6.07 140.8
15.3 78 94.2 9.31E+06 8.81E+09 3.82 4.80 6.51 143.6
0 74 91 8.08E+06 7.72E+09 3.12 3.87 5.11 134.1
5.7 74 91.4 8.54E+06 8.11E+09 3.30 4.11 5.48 136.8
9.5 74 91.7 8.85E+06 8.42E+09 3.44 4.31 5.78 138.9
15.3 74 92 9.31E+06 8.81E+09 3.63 4.56 6.18 141.5
15.3 65 87 9.31E+06 8.81E+09 3.23 4.04 5.49 135.1

The above performance estimates would be repeated for the range of uprate conditions that need
to be considered to further define the business case economics. But a few other aspects of the
power uprate evaluation need to be addressed to determine the economic value. Condensate
temperatures at the uprate conditions should be evaluated to ensure no limiting condensate
polishing resin temperatures are exceeded. Multipressure condensers also require the hotwell
reheat temperatures to be predicted during the uprate. This can be accomplished by assuming
caloric mixing of the condensate from all compartments and then a full flow reheating to
reasonably approach the high-pressure compartment steam saturation temperature.

1-9
In extreme cases, the capacity of existing drain trays to handle the extra condensate should be
investigated. The condenser non-condensable gases and/or air inleakage after an uprate should be
essentially unchanged except in the case of a Boiling Water Reactor (BWR) facility. The
production of non-condensable radiological gases increases in proportion to the associated uprate
level of a BWR facility. The effect of this increase in non-condensables may require further
evaluation. The overall duty of any steam jet or vacuum pump air removal equipment is
increased while at the same time the condensate temperature used to condense the steam in a
steam jet intercondenser also increases. Similarly, the condensate that seals the vacuum pump
may also see a significant temperature increase that can limit or reduce vacuum pump
performance.
When an even more accurate estimate of the condensers thermal performance for a significant
level of power uprate is necessary, a special computational fluid dynamic (CFD) analysis of a
condenser can provide the solution. This method has been available since the 70s but computer
core space, matrix solution speeds, input-output limitations and virtual modeling inflexibility
restricted its use and effectiveness to only long-term research projects. Now with the
revolutionary improvement in computer storage, graphics and processing speed, this technology
has enjoyed a wider and more practical application with condensers. The equations of motion of
the steam, the condensing heat transfer and mass transfer based on the conservation of mass,
momentum and energy are applied separately to each of thousands of small elements that
simulate a virtual model of the condenser. Then, the composite interaction and response is
determined by iteration to satisfy the boundary conditions of the cooling water flows, cooling
inlet and outlet temperatures, overall heat loads, etc.
One example of a recent CFD application was to a more than 60,000 tube, three-pressure
condenser in a nuclear plant that would be operating at a power uprate of over 25% more than
the design of the original condenser tube bundle. In this case, a model of over 350,000 elemental
three-dimensional volumes was developed that duplicated one of the two symmetrical tube
bundles and the corresponding local steam space geometry. After the model geometry was
verified by comparisons of the model to detailed condenser design drawings and the tube bundle
patterns, computer runs that captured key operating conditions involving appreciable steam side
effects were performed to validate the model. Verification encompassed the unique absence of
any tube vibration history in the operation of this condenser for over 20 years. Then, confident
predictions were made of the power uprate turbine exhaust pressures at peak conditions in the
summer to estimate the response of the steam turbine and its generation at this evaluation point.
Tube Vibration
Condenser support plate spans are selected during their design to avoid tube vibration. Besides
the design flow velocity, the span selected by the designer is basically dependent on the lowest
condenser pressure, the tube material, its diameter and wall thickness. The higher flow rates
associated with a power uprate increase the potential for damaging flow-induced vibration. Since
a power uprate has the potential for causing tube vibration, it is prudent to determine the location
and length of anti-vibration stakes. These stakes are typically inserted perpendicular to the
bundle at the center of a support bay span to arrest any tube vibration. The uprate impact must be
quantitatively considered in order to determine if, how long and where anti-vibration stakes need
to be located in a condenser tube bundle.

1-10
Condensers have experienced damaging tube vibration since the mid-1960s. The damage is
either from mid-span collision, fretting or fatigue failure at the point of support. Heat exchanger
manufacturers generally accept the fluid-elastic, flow induced vibration mechanism as the root
cause. One numerical tool that allows this evaluation is the Connors Criteria. It can be applied to
separate stable and unstable fluid-elastic vibrations and is graphically evident on the plotted
results of those calculations. Note that this type of fluid-elastic vibration of condenser tubing has
been observed during simulations in controlled laboratory conditions many decades ago.
An adaptation of the Connors fluid-elastic method to condensers allows an estimate of the
potential for damaging tube vibration during a power uprate. This quantitative method of
estimating the onset of damaging condenser tube vibration captures the major aspects of the
actual physics of the fluid-elastic aerodynamics. It requires an extensive calculation effort
accompanied by a significant amount of detailed condenser design and tube bundle data. During
full uprate load operation at the stations historically low backpressure, it is necessary to estimate
the axial and circumferential steam flows around the entire tube bundle. In order to approximate
the steam velocity throughout the bundle, it is essential to also determine the condensation of that
steam as it enters and travels through the bundle toward the air cooler. The velocity estimate
requires measuring the detailed bundle pattern geometry on a large scale drawing of the tube
sheet to establish flow areas at several planes on the periphery and in the interior of the bundle.
The condensation heat transfer and the steam flow are estimated through sections of the
longitudinal and radial portions of the bundle in order to define the local steam velocities across
the tubing. Finally, based on the mechanical properties of the tubing and the existing support
plate spacing, the unit weight, damping and natural frequency is estimated. These characteristics
are compared to determine if, when and how deep into the bundle damaging fluid-elastic tube
vibration is likely during a power uprate. The resulting information establishes the length and
location of the anti-vibration stakes required to avoid tube vibration. The lengths, numbers and
locations can be used as the basis of a subsequent condenser tube staking specification.
An example of the results of tube vibration estimates for a power uprate condition follows. The
inlet temperature to the condenser was first determined as the lowest historical temperature that
had occurred in the history of the plant. An estimate of the condenser backpressure at the power
uprate condition was made first on the basis of that lowest inlet water temperature, the CW pump
flows, the uprate heat duty and the apparent condenser cleanliness. The pressure will be found to
be higher than the existing power level condenser pressure because of the increased duty and the
possible impact of the performance of any cooling tower during an uprate.

As stated previously, the Connors criteria separate a small amplitude, stable vibration from
unstable, large amplitude fluid elastic tube vibrations. This criteria specifically states that
vibration damage will not occur with conditions established by the following equation:

2
D
W
D f
V
o
tot
n

The left side of the equation is defined as the Vibration Parameter and the right side of the
equation is defined as the Damping Parameter.

1-11
Where:
= V Local steam velocity
=
n
f Tube natural frequency
= D Tube O.D.
=
o
Steam density
=
tot
W Tube weight
= Damping factor
= Factor describing tube array, pitch, diameter and configuration
Table 2 and Figure 1 show typical results of this type of analysis. These results are for the low-
pressure shell of a three pressure, multipressure condenser. The shell contains 30,732 1 in. dia.
22 BWG type 304 stainless steel tubes of ~27 feet in length. Bay 1 is at the inlet end of the low
pressure (LP) condenser compartment and Bay 6 at the warm or outlet end. Note that the
representative table only covers the tube bundle peripheral area, as it would have been excessive
to include ten pages of tables for each distance into the bundle. However, the plot includes
depths up to 5 feet deep into the bundle.

Table 2
Tube Bundle Peripheral Flow Area ONLY

Section Bay 1 Bay 2 Bay 3 Bay4 Bay 5 Bay 6 Bay 7
Length (Ft.) 4.75 8.53 12.14 15.92 19.54 23.32 26.93
Temp Out (
o
F) 46.74 48.08 49.33 50.59 51.77 52.95 54.07
Temp Rise (
o
F) 1.74 3.08 4.33 5.59 6.77 7.96 9.07
% Steam Cond. 19 34 48 62 75 88 100
Steam Flow (Lb/Hr.) 167216.52 128766.9 119519.2 121342.6 112600.5 114347.2 106109.1
Flow Area (Sq Ft.) 53.958 43 41.04 43 41.04 43 41.04
Sp. Vol 373.8 373.8 373.8 373.8 373.8 373.8 373.8
Velocity (Ft/Sec) 321 310 302 293 284 276 268
Velocity Streak
(2xVelocity)
642 620 604 586 568 552 536
Full V / f
n
x D 185 113 101 107 95 101 89
Half V / f
n
x D 46 28 25 26 23 25 22
W
tot
x g/rD
2
1169 1169 1169 1169 1169 1169 1169

1-12

Figure 1
Critical Velocity versus Damping Parameter

The results shown in the plot indicated that the peripheral tubes in all the condenser bays were
prone to damaging fluid elastic vibration. In Bay 1, the extreme cold end of the condenser, such
damage is possible down to a depth of almost 5 feet. This analysis is a useful tool when
considering the potential of tube vibration damage resulting from the increased steam flow
required for a power uprate.
It should be appreciated that the Heat Exchange Institute has published an algorithm within their
standard that determines the condenser tube support spacing that avoids severe vibration. It is a
rule-of-thumb procedure based on an empirical static load deflection concept. The HEI method
does not reflect the actual physical mechanism of the tube vibration as a fluid-elastic whirling
vibration. In practice, that calculation has been found to provide a conservative estimate of the
required support plate spacing for new condensers. And because the HEI method does not
address the actual physics, it cannot effectively be used to determine the length or location of

1-13
anti-vibration stakes for an existing condenser that has to be redesigned to accommodate the
higher condensing steam flows of a power uprate.
Cooling Tower Evaluations
Many power stations, particularly the larger generators and nuclear plants, use closed-cycle
cooling systems that are served by wet cooling towers. A power uprate at such a plant will cause
a direct increase in the evaporation and higher return water temperatures to the condenser. Often,
the additional makeup flows are required to offset evaporation increases. These may be limited
by environmental regulations.
In order to estimate the increased water required for the makeup, it is first necessary to know the
number of cycles of concentration of the circulating water. During warm weather conditions
characterized by high wet bulb temperatures, it has been empirically determined that the
evaporation component of a wet cooling tower is no more than 80% of the total heat duty [5].
This rule of thumb applies equally well to mechanical or natural draft towers, as well as cross or
counter-flow designs. Since each pound of evaporating water releases about 1000 BTUs, an
approximation of the added evaporation as a result of the power uprate can be assessed. Of
course, more exact estimates of the evaporation can be made. For the power uprate condition,
these may be accomplished by determining the quantity of moisture picked up by the cooling air
as its state changes from the entering air temperatures to the usual saturated exit air condition.
The added cooling tower blowdown is estimated by dividing the extra evaporation by one less
than the number of concentrations that currently are employed. Added plant cooling tower
makeup is the total of these two elements. Note, that in general, drift losses are minimal and do
not need to be seriously evaluated in the estimate of additional makeup flows.
The change in performance of the cooling tower due to a power uprate can be estimated from the
existing manufacturers performance curves by using the uprate cooling range and reading the
curves to evaluate the increase in approach to the wet bulb temperature. That provides the inlet
water temperature to the condenser. It is important that these curves be adjusted to be compatible
with the results of any recent performance tests, otherwise a significant understatement could
result. The actual CW pump flow measured by an appropriate method should be used to modify
the curves. To make the corrections to the existing cooling tower performance based on test
results and or changes in the CW flow from design, new curves must be drawn accounting for
the current performance non-dimensionally with the actual CW flow and the fill characteristic vs.
the water to air ratio. The detailed method to apply these adjustments are beyond the scope of the
paper but can be derived from the information shown in References [3,5].
Cooling System Upgrades
Finally, it should be noted that if the existing cooling system equipment performance is found to
be marginal such that a station power uprate would have limited benefits, it may be cost effective
to consider incorporating a modification to the cooling system in the uprate plans. Below is a
brief list that identifies some of the typical modifications that can be considered:

1-14

Upgrade Cooling Tower Fill
Upgrade Cooling Tower Fans or Airflow pattern
Upgrade CW Pump Performance
Install a modern Condenser Tube Ball Cleaning System
Improve Condenser Performance
Lag Unlagged Heaters and Extraction Lines
Replace Condenser with Modular Tube Bundles.
Retube Condenser with optimum Tube Material & Gauge
Install a Helper Cooling Tower

The cost(s) associated with each enhancement option must be carefully weighed against the
potential to increase plant performance and its corresponding economic benefit. Obviously, these
types of modifications represent a major change in the uprate project scope and costs, but may be
the only way to permanently change a pigs ear into a silk purse.
Conclusions
The power uprate constitutes a method of cost-effectively increasing the generation of existing
nuclear and fossil power plants. Significantly more waste heat will accompany the power uprate
and must be accommodated by the existing cooling system. When considering these projects, it
is important to review the current condenser and cooling tower performance capability to
determine the degree to which that equipment will be compatible with the uprate during peak
demand conditions. The paper describes the method of investigating and quantifying the current
performance of the cooling system & its components. Then, using these baseline performance
estimates one can extrapolate the uprate capability. Of the equipment involved, the condenser is
often the most sensitive limiting the benefits of a future uprate.
It was concluded that the HEI condenser performance predictions do not accurately take into
account the steamside effects from the added turbine exhaust steam flow that accompanies a
substantial uprate. The HEI method often seriously underestimates the resulting backpressures.
The paper recommends applying an extension of the resistance heat transfer method to provide a
more reasonable estimate of the backpressure and condensate temperatures since it accounts for
steamside effects.
Other uprate impacts discussed include the increased potential for damaging condenser tube
vibration especially during winter operation at low operating backpressures. The vibration
usually causes failure at the midspan or due to wear/fretting at the supports. The Connors
technical criteria was introduced in order to determine the location(s) and depth into the tube
bundle for installation of anti-vibration stakes to prevent damaging fluid-elastic vibration.

1-15
The paper also determined that a power uprate study must address the post uprate capability of
the cooling tower. The tower will produce higher condenser return temperatures, adding to the
turbine exhaust pressure. A power uprate will require increased cooling tower evaporation,
blowdown and makeup. These may be subject to heightened environmental regulations. The
paper recommends and discusses applicable methods to estimate the power uprate effects on the
cooling tower performance.
Finally, the paper indicates that if the current cooling system equipment performance is marginal
allowing only a limited station power uprate, it may be cost effective to consider incorporating
system improvements. For each improvement option, a detailed feasibility and cost study is
necessary including the cost of the uprate modification and the present value of the annual
increased station revenues. Some of these modifications are listed in the paper and may include
upgrade of cooling tower fill, a tube ball cleaning system, higher capability CW pumps, modular
condenser replacement, a helper tower, or other modifications to the CWS equipment.
Taking into account all the power uprate impacts on the system performance including the
addition of any cooling system modifications will appreciably contribute to the success of the
overall project.
References
1. Burns, J.M.; Almquist, C.; Hernandez, E.; Tsou, J.; "Accurate Condenser Performance
Monitoring Guidelines Provided by New ASME Condenser Test Code", EPRI Heat Rate
Improvement Conference, May 1994.
2. ASME PTC 12.2, Performance Test Code on Steam Surface Condensers, 1998.
3. ASME PTC 23, Performance Test Code on Atmospheric Water Cooling Equipment, 1986 or
2002.
4. Heat Exchange Institute, Standards for Steam Surface Condensers, Ninth Edition, 1995.
5. Burns, J.M. and Brocard, D.; "Cooling Towers," Section 4.8. Handbook of Energy Systems
Engineering: Reference Text, J. Wiley & Sons, N.Y., 1985.

1-17
CFD ANALYSIS PREDICTS CONDENSER PERFORMANCE AFTER LARGE
POWER UPRATE OF THE QUAD-CITIES AND DRESDEN STATIONS
N. Rhodes
Consulting Engineer
New York
C. D. Hardy
Heat Exchanger Systems Inc.
Weymouth, Massachusetts
J. M. Burns
Topsfield, Massachusetts
T. B. Madden
Stone & Webster Inc.
Cherry Hill, New Jersey

Abstract
This paper presents the results of work carried out to study the Extended Power Uprate (EPU)
operating conditions for Exelons Dresden Nuclear Power Station Units 2 and 3 and Quad-Cities
Nuclear Power Station Units 1 and 2 condensers. The study involved the development of a three-
dimensional condenser simulation model based on computational fluid dynamics (CFD)
techniques. Subsequently, that model was applied to predict the turbine exhaust pressures for
the cumulative 36% increase in design condenser steam flow that was associated with EPU
conditions.
The CFD model reflected the major design details of the condenser including the tube bundle
pattern of this single pass, three pressure multipressure configuration that contains an
intermediate waterbox . To validate the accuracy of the model, the condenser performance was
first evaluated at current conditions using the existing steam mass flow, cooling water flow rate
and temperature for several conditions measured at the plant. Comparisons of the CFD
simulation to actual observations verified that the model results closely predicted the
performance and physical response of the condenser.
Following this verification, the model was then applied to ensure the business case that justified
the costs of the plant uprate would be realized. Specifically, the EPU 36% increased design
steam flow condition was combined with summer water temperatures and simulated by the CFD
model to be reasonably certain the turbine exhaust pressures would not exceed manufacturers
limits and at the same time, that the target EPU plant generation of 912 MW would occur.

1-18
Introduction
This paper describes work carried out to study the Extended Power Uprate (EPU) operating
conditions for Exelons Dresden Nuclear Power Station Units 2 and 3 and Quad-Cities Nuclear
Power Station Units 1 and 2 condensers. The ultimate objective of the condenser performance
study was to ensure the business case that justified the costs of the plant uprate to 912 MW
would be realized during warm summer conditions. In comparison to the design basis of the
existing tube bundles, EPU operation would cause 36% more steam flow to be condensed.
Because of the appreciable condenser steam-side heat transfer, condensate and flow effects that
would be associated with this level of extra condenser steam flow, it was determined that the
typical HEI (3) condenser performance algorithm would not be applicable since HEI only
addresses water-side tube bundle effects. This engineering study involved the development of a
three-dimensional condenser simulation model based on computational fluid dynamics (CFD)
techniques. Subsequently, that model was applied to predict the turbine exhaust pressures for
the appreciable 36% increase in design condenser steam flow.
The CFD method separates a condenser steam space into small connected discrete volumes
called cells that can be characterized mathematically. After the geometry is modelled and the
three dimensional cell mesh created, all the equations that apply are simultaneously solved to
each of the volumes to describe the detailed physics that is occurring at that cell. These
equations capture the fluid motions of the steam, the condensing heat transfer and mass transfer
based on the conservation of mass, momentum and energy, and account for parameters such as
gas concentration, heat transfer resistances, transport properties and condensate inundation. The
interactions and gross response of all these small volumes are then solved iteratively to estimate
the conditions and compatibilities at the boundaries of the model, i.e., the cooling water flows
and inlet temperature, , overall heat loads, turbine exhausts and non-condensible-vapor off-take.
The techniques used in this particular program have been refined and applied specifically to
steam condensers since the initial general purpose CFD method was first utilized for condensers
by the CEGB in the early 1980s.
This CFD model reflected the major design details of the condenser including the tube bundle
pattern of this single pass, three pressure multipressure configuration that contains an
intermediate waterbox . To validate the accuracy of the model and technique, the condenser
performance was first evaluated at current conditions using the existing steam mass flow, cooling
water flow rate and temperature for several conditions measured at the plant. Comparisons of the
CFD simulation to actual observations verified that the model results closely predicted the
performance and physical response of the condenser. Following the verification, the model was
applied for the EPU conditions.
Following EPU, the Dresden and Quad-Cities condensers will be required to condense
approximately 8 million lbm/hr, thus stretching the condenser capability to accommodate turbine
exhaust steam flow by over 35%. The use of a sophisticated three-dimensional model as applied
in this study assists in predicting the increased effects of steam side pressure drop associated
with higher steam velocities and the effect of additional condensation which increases
condensate inundation.

1-19
Physical Situation
The Dresden and Quad-Cities condensers are identical in design. The main surface condenser
unit consists of two tandem tube bundles with an intermediate waterbox within a single,
rectangular shaped shell that is oriented parallel to the turbine shaft. It is a three pressure multi-
pressure, single-pass, vertically divided design that was originally designed by Ingersoll-Rand in
the late 60s and early 70s to serve a 715 MW General Electric Company turbine generator.
The condenser design parameters are listed in Table 1 and the tube bundle configuration is
shown in Figure 1. Nine main support plates in each of the three pressure zones extend from the
top to the bottom of the tube bundle, perpendicular to the tubes and are welded to the shell. There
are two support/partition plates, one in each bundle, that run from the top of the condenser to a
water seal below the tube bundles, and are uninterrupted running across the condenser. These
partition plates divide the condenser into the three separate pressure compartments. The lowest
pressure compartment is at the circulating water inlet end of the condenser, the highest pressure
is at the outlet end and the middle compartment is the intermediate pressure compartment.
The cooling water passes through the low pressure section, partially through the intermediate
vacuum section and then it enters an intermediate waterbox. The tubes are continued through the
intermediate pressure section, into the high pressure section and out of the condenser.
In cross-section, the condenser is symmetrical about its vertical centre-line and so only one half
of the condenser has been modeled. Figure 1 shows this half-section with the position of the tube
bundle superimposed. Figure 2 illustrates the complex geometry of the internal structure. This
figure shows a perspective detail of the intermediate pressure section, including the waterbox,
tube support plates and the dividing plates between the other pressure sections as well as the
plates which control the movement of the uncondensed steam/air mixture towards the air offtake
section in the low pressure compartment.
The condenser contains a total of 61,464 tubes divided into two tube bundles. (Note that all
quantities are referred to the whole condenser and not the half-section modeled). Figure 1 shows
the distribution of the tubes in one bundle. It can be seen that the upper half of the tube-nest is
characterized by raysof tubes aligned almost vertically, with two triangular sections on either
side of the lower part of the upper section. Although not shown, the outer edge of these latter
sections have horizontal lanes, created by the removal of one row of tubes in every four rows.
These extend for about ten to twelve tube rows.
In the lower part of the tube nest the outer parts all have steam lanes similar to those described
above. The horizontal lanes in this region extend from the outer perimeter to the fully-tubed
sections on either side of the centre-line of the tube nest.
Non-condensable gases are removed at the cooling water inlet end. Air removal is by way of a
series of apertures in the tube support plates which direct the non-condensables to the cooling
water inlet end. The air-cooling section is bounded by a divided horizontal tray below the upper
half of the tube bundle, and sloping trays below. The shorter sloping tray within this section
corresponds with apertures in the tube support plates which alternate above and below these
plates along the length of the condenser. These direct the non-condensable gases and
uncondensed steam through the small tube bundles within the air cooling section as it flows

1-20
towards the outlet. A second divided horizontal tray is located below the air-cooling section
between the two sections of the lower bundle.
The design data which has been used in the study is given in the following table:
Table 1
Condenser Parameters
Design Variable Value
No. of Tubes 61,464
Tube Length 40ft 4.75in
Total Surface Area [ft
2
] 650,000
Tube Outside Diameter [in] 1.0
Tube Wall Thickness 22 BWG
Tube Material ASTM 304 SS

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,, support plates, and 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 References 1 and 2.
The condenser model utilizes a general-purpose Computational Fluid Dynamics (CFD) program,
the CFX code developed by AEA Technology, in conjunction with a highly developed model of
the heat and mass transfer processes. The finest distribution of cells is in the cross section of the
condenser. In the present study a grid of 61 by 117 cells divide the condenser in the horizontal,
x-direction, and the vertical, y-direction, respectively. Longitudinally, there are two divisions
between each support plate, resulting in 50 cells in the z-direction. Thus, the condenser model
utilizes a grid of 356,850 cells.
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 into the condenser. The velocity is not uniform, but varies with
position at the inlet. The variation in velocity reflects the Hertzog Hood design and was
included in the modelling.
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, in this case steam-jet air ejectors, are not modelled
explicitly.

1-21
Table 2 presents the condenser parameters used for the base case, Case 1 and the EPU case, Case
2. Note that where applicable the values are halved for the model as only one half of the unit is
simulated.
Table 2
Operating Parameters
Case 1 Case 2

Average CW inlet temp (F) 34.5 90
Circulating water flow (gpm) 364,000 473,000

Main steam heat load (BTU/hr) 5.68464E+09 6.78422E+09
Main steam flow (lbm/hr) 6,324,000 8,125,755
Main steam enthalpy (BTU/lb) 934.4 944.3

Cold Hood Condenser Duty (BTU/hr) 1.91064E+09 2.33745E+09
Intermediate Hood condenser duty (BTU/hr) 1.85642E+09 2.28755E+09
Hot Hood condenser duty (BTU/hr) 1.82798E+09 2.24935E+09

Cold Hood Condenser steam flow (lbm/hr) 2,108,834 2,708,585
Intermediate Hood condenser steam flow (lbm/hr) 2,108,558 2,708,585
Hot Hood condenser steam flow (lbm/hr) 2,108,325 2,708,585

Cold Hood enthalpy (BTU/lb) 929.3 936.4
Intermediate Hood enthalpy (BTU/lb) 929.4 943.1
Hot Hood enthalpy (BTU/lb) 929.5 953.4

Air inleakage (lbm/hr) 90 (actual) 211.1
Hydrogen carryover (lbm/hr) 20.9 37
Oxygen carryover (lbm/hr) 153.8 295

Tubes plugged 450 3,518
Tube cleanliness 100% 80%

Presentation of Results
Typical flow visualization plots showing velocity vectors and steam concentration are given in
Figures 3 and 4. The velocity vectors show the magnitude and direction of the steam/air flow.
These are made at a two-dimensional x-y plane within the low pressure vacuum section.
Figure 3 shows the velocity distribution over the whole section. The main features which can be
seen are:
The variation of velocity across the top of the tube nest in the upper part of the figure. It is
interesting to note that this maldistribution of flow persists down to the top of the tube nest,
and some flow recirculation is noticeable due to the outward sloping shell.

1-22
The frictional resistance of the upper tube nest causes part of the flow to divert around the
top of the nest and flow down both sides of the bundle in the outer and central lanes.
Where the horizontal drain tray protrudes outside the bundle, at the top of the air-cooling
section, the flow accelerates slightly.
The flow down the central lane continues around and underneath the bundle and causes the
flow on the outer lane to be unsymmetrical. The steam enters the lower bundle on the outer
side about half-way up the lowest section, the flow stagnating in the outer lane in this region.
The velocities within the tubed regions are generally lower due to condensation and frictional
resistance.
Figure 4 shows the steam/air concentration at this plane. In the upper part of the bundle, low
concentration regions can be seen on either side. In the lower bundle, similar features can be
seen on the right-hand side below the air-cooling tray and the lowest rain tray. These correspond
with low velocity regions of the flow, and there is little pressure gradient to drive the flow
towards the air-cooling section, in contrast with the left side, where the flow continues through
the denser inner tubed region. Note that the steam/air concentration is assumed to be uniform at
the inlet and is determined from the respective steam and non-condensable flow rates. As
condensation proceeds, the relative concentration of air increases as can be seen in the figure.
Case 1: Model Verification and Validation
All CFD models should undergo a verification and validation test to establish their level of
accuracy. In this case, its verification determined that the computer model accurately
represented the tube bundle, flow conditions and geometry of the condenser. That task involved
a comparison of the model to numerous design drawings and to the plant CW flow and
temperature data. Next, the validation process settled the extent to which the model and its
simulation reflected the real world from the perspective of its intended use (performance during
operation). The model was subsequently used to predict the Case 1 winter full load conditions
and by a visual and numerical examination of the steam flow patterns and velocities at those low
backpressures, estimate if the CFD model would have predicted condenser tube vibration which
had been experienced under certain conditions.
Two different types of validation tests were applied. The first in this instance was a prediction of
representative, existing full load operation at winter conditions that cause a low condensing
pressure. The results of this estimate compared well against the plant operating data as Table 3
attests. Therein, the average compartment condenser pressures during winter full load operation
at the turbine-condenser flange of the Quad-Cities condenser are compared with measured data.
Table 3
Comparison of Measured and Predicted Condenser Pressure for Case 1
cold hood intermediate hood hot hood
measured 1.11 1.3 1.53
predicted 1.12 1.4 1.71

1-23
The predicted pressures are averaged across each inlet section at the turbine-condenser flange. It
can be seen that the predictions are in reasonably close agreement with the data, although there is
a tendency to slightly over-predict the pressures at the high vacuum pressure inlet, the difference
being 0.18 in Hg.
Because of the high steam velocities that occur during low pressure winter operation, tube
vibration can occur to condensers with tube support spans that are too long. Tube vibration
damage had frequently been experienced by many condensers fabricated with the relatively long
support spans that were the norm for the time frame of the design of Quad-Cities and Dresden.
But no tube vibration damage had been observed at Quad-Cities until the early 90s despite over
20 years of operation. Thus, to establish the validity of the model, advantage of that fact was
utilized in an examination of the detailed steam velocity levels of the low pressure compartment
in the simulation of Quad-Cities winter operation.
That study provided a unique and powerful indicator of the validity of this model because the
CFD method simulates (also) the condenser steam-side effects and parameters during operation
and these are the ones that were to be significantly altered by EPU. The CFD validity was
established after the winter full load condenser pressure level was predicted. Then the (non) lack
of appreciable tube vibration was determined from applying the Connors Criteria to the detailed
CFD steam velocity results at various cross sectional tube bundle planes in the low pressure
compartment. All steam velocities tested were essentially below the Connors threshold.
Figure 3 illustrates the predicted velocity of the steam around the tube nest. The velocity vectors
indicate the magnitude and direction and are plotted for each grid cell. Numerical data from he
model was provided as u- and v-component velocities within and around the tube nest to
facilitate the Connors analysis.
A further performance run was conducted for summer conditions at the existing full load prior to
carrying out the EPU case. As well as providing additional validation of the model, the results
provide a basis for comparison with the EPU case which is also for a summer condition. The
plant-measured conditions for this run were as follows:
Table 4
Full Load Conditions for Summer Case
Inlet CW temperature (F) 90
Main Steam Flow (lbm/hr) 6,600,000
Main Steam Heat Load (BTU/hr) 5.67E+09
Tubes plugged 3193
Equivalent HEI Cleanliness 0.76

Table 5
Comparison of Measured and Predicted Condenser Pressure for 90F inlet temperature
cold hood intermediate hood hot hood
Measured 3.07 3.7 4.53
predicted 3.2 3.69 4.58

The results in Table 5 show reasonably good agreement with measurements.

1-24
Case 2: Power Uprate Conditions
The predicted pressures at the EPU conditions are presented in Table 6.
Table 6
Turbine exhaust inlet pressures for EPU conditions
Offtake pressure cold hood intermediate hood hot hood steam/air ratio
at outlet
3 4.22 4.806 6.06 5.37
1.79 3.78 4.44 5.63 10.1

The offtake pressure for this simulation was set at values of 1.79 and 3.0 in Hg.. At this
condition, the steam/air ratio at offtake was predicted to be 10.1 and 5.37 respectively, which is
to be compared with the value of 5.5 obtained for the full load case at 90 F inlet temperature. If
the SJAEs could not handle this ratio then the condenser pressure would adjust itself until a
reasonable operating ratio was achieved. Consideration therefore may need to be given as to the
consequences on the air offtake equipment of higher flow rates.
Finally, it should be borne in mind that the exceptional detail of the CFD model was designed to
simulate the physics of the condensation heat transfer and compressible fluid flow that occurs
within the tube bundles of an actual condenser. The CFD model faithfully duplicated the
geometry of the space between the tube bundles and turbinecondenser flange. Though it
included a representative turbine steam velocity profile as well as the two large heaters, the
extensive structural bracing, supports and extraction lines were not incorporated. Hence, other
than the minimal effects of wall friction, the CFD model did not capture the exhaust steam
pressure loss between the end of the turbine skirt and the condenser tube bundles. Instead, past
engineering experience was used to approximate that pressure loss. For the LP condenser zone,
it is estimated that a 0.15 in Hg loss between the CFD model result and the turbine flange would
occur during the EPU Case conditions. Then based on the comparative steam specific volumes,
that pressure loss value should be modified to 0.13 and 0.10 in Hg to respectively represent the
IP and HP condenser zones. These pressure drops should be directly added to the CFD model
tube bundle results in order to estimate the expected EPU turbine exhaust pressure conditions.
Conclusions
This paper describes the application of CFD-based condenser model to predict the three-
dimensional flow and heat transfer behaviour in the Dresden and Quad-Cities condensers. The
method was utilized because the costs of the EPU were to be justified by confirming that
912MW of generation could be obtained during warm summer conditions while turbine exhaust
pressure limits would not be exceeded. For this target condition, the condenser design steam
flow will increase 36% after EPU and this exacerbated steam side condition is not captured by
the HEI condenser performance algorithm. Validation of the model accuracy was first
demonstrated by comparing its predicted performance during existing operation and its
prediction that was compatible with the physical reality that there would be no extensive tube
vibration at this condenser despite its many years of cold weather, full load operation at low
turbine backpressures. Following the model development and validation phase, the model was
then applied to EPU conditions.

1-25
Mainly, the results of this CFD study established that during summer full load EPU conditions
912 MW will be obtained and turbine exhaust pressures will be slightly below the
manufacturers limits. The comparison of the CFD model results vs. plant operating data shows
that it accurately predicted the exhaust pressure measured at the plant and had provided a reliable
early confirmation of the expectation of producing 912MW.
In addition, the results of the study provided an insight into the flow behaviour within the
condenser. With regard to the flow, the asymmetrical behaviour in the lower part of the bundle,
caused by a combination of the flow maldistribution, the expansion in the side of the shell and
the upper horizontal rain tray gave rise to poorer performance in the right part of the tube bundle.
Future applications of the model could include a determination of how to operate the circulating
water system to avoid vibration induced tube failures, specification of optimal staking locations,
and to develop modifications to improve performance.
References
1. Al-Sanea S, Rhodes N, Tatchell DG & Wilkinson T S A computer model for Detailed
calculations of the Flow in Power Station Condensers.Proc. Condensers: Theory and
Practice, IChemE Symposium, Series No. 75 pp78-80 (1983)
2. Al-Sanea S A, Rhodes N and Wilkinson T S (1985) "Mathematical Modelling of Two-Phase
Condenser Flows" Presented at 2
nd
International conference on Multi-Phase Flow, London
3. Heat Exchange Institute Standards for Steam Surface Condensers, Ninth ed., 1995.

1-26

Figure 1
Condenser Cross Section

Figure 2
Perspective view showing intermediate waterbox and detail of internal structure

1-27

Figure 3
Velocity Vectors at the mid-plane of the LP Section

.

Figure 4
Steam Concentration Contours

1-29
SURFACE CONDENSERS, STEAM DUMP, & DESIGN
RELIABILITY CONSIDERATIONS
David H. Cooley
Alstom Power
Fossil power plants typically used steam dump for up to 10% - 20% load on start up (one start
every year to year and a half) and under turbine trip conditions that normally had a fast decay
from 110% load in 30 min. to 1 hour. The typical fossil power plant had a steam dome that
contained a feedwater heater and extraction piping. The height of the steam dome was, more
often than not, set by these requirements and/or keeping the waterboxes under a mezzanine floor.
Thus steam domes usually had an average height of 12 to 16 feet with some large MW plants
reaching 30 feet.
These type of designs provided the volume necessary to dissipate the energy of the steam dump,
slow velocities to acceptable levels and locate the dump tubes far enough from the tube bundles
to prevent localized high velocities or direct impingement on the tubes. There were few growing
pains in the late 1960s and early 1970s that were primarily centered around proper draining of
steam dump lines coming to the condenser and the philosophical question of proper location of
dump tubes within the condenser to protect the turbine. After these were resolved, the industry
proceeded with virtually no problems associated with this service.
In the mid 1980s the industry evolved from fossil power plants to combined cycle power plants.
Combined cycle power plants brought new modes of operation that include multiple starts per
year, multiple HRSGs 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 change in the type of power plant coupled with the privatization of the Utility industry
changed the financial basis for power plants and hence the system designers considerations on
space and components. These system design considerations have included elimination of
diverter dampers thus requiring continuous steam dump, thinner wall tubing, relocation of the
dump valves closer to the condenser, introduction of enthalpy control systems for
desuperheating, and less available space for the condenser.
The combined cycle power plants have quickly grown in size and complexity in the last 10 years.
Plants of 600MW total combined output have become commonplace. Plant designs are ever
evolving and single shaft units where the gas and steam turbine are on the same shaft are just
recently being brought on line. Hence, plant start up requirements have been continuously
evolving and therefore the requirements on the condenser have changed considerably. Yet the
plant designers have typically been unable provided the condenser designer with the information
related to the start up requirements and operation modes that could affect the condenser design.
The net result is that the industry has recently been experiencing condenser failures that have
been associated with steam dump operation. These failures have included fatigued tubes and
eroded tubes. When these failures occur at initial by-pass operation, the start up is delayed.

1-30
Repairs may be costly and time consuming and other equipment in the plant may have been
affected. Thus the successful operation of the condenser in the steam dump mode is critical to a
successful plant start up and long term operation and as such deserves the proper attention in the
design phase.
Steam dump operation has generally not had the focus necessary for the industry to understand
that this mode of operation is severe. In combined cycle power plants it is often the criteria that
drives the condenser design. When steam dump flows exceed approximately 120% - 140% of
normal design flows the steam flow areas within the condenser are often dictated by the steam
dump mode of operation
Steam entering the condenser from the turbine has already been expanded to the saturation
conditions. It is delivered to the condenser over a large area at relatively benign velocities and
with velocity vectors directed toward the condenser tube bundles. Steam dump generally enters
the condenser from the dump tube superheated about 300F, over a relatively small area at sonic
velocity and with velocity vectors generally perpendicular to the normal steam flow direction.
The condenser is required to accept this flow and desuperheat it to saturation conditions, reduce
the velocities to acceptable levels and provide proper distribution over the length of the tube
bundle. At the same time the turbine and condenser structure and tube bundles must be protected
from overheating, erosion and localized damaging high velocities.
There are many different types of steam dump devices (i.e. perforated dump tubes, reverse
conical dispursors, multiple expansion device, etc) used distribute the steam within the
condenser. With any of these steam dump devices the reliability of the condenser operation
under steam dump can be assured with the following three criteria:
Plant designers must provide the condenser designer with the start up scenarios. With
multiple HRSGS different steam dumps can be operating at different times and often partial
operation can produce distribution velocities in the condenser that may be unacceptable to
the condenser designer.
Steam delivered to the condenser must be dry at all points of operation (25F to 75F
superheat). Erosion of condenser parts and tubes can happen in minutes depending upon the
wetness of the steam.
Plant designers must provide space for adequate volume in the steam dome and distance from
the condenser tube bundles to allow for effective energy dispersion and velocity reduction of
the dump steam prior to entry into the tube bundle. This prevents localized high velocities
and/or heating that can cause tube fatigue failures or overstressing.
Many plants have multiple combustion turbines and multiple HRSGs. The start up and
operational scenarios are many. The condenser can see multiple combinations of steam dump
flows over the entire range of cooling water temperatures. Consequently a condenser designed
for 2.00 in. HgA can typically see pressures ranging from 0.8 in. HgA to 5.00 in. HgA. This is a
5.8 fold variation of the specific volume causing a comparable variation in velocities over the
range of operation.

1-31
Placement and/or interconnection of steam dump piping can often mean the difference between
damaging localized high velocities and safe operation at off design situations. This is
particularly important in side exhaust arrangements where interconnection of all steam dump
sources outside the condenser to allow equal flows to both condenser shells at all times is
essential to preventing cross flow through the turbine and potential turbine damage.
Down flow and axial exhaust turbine suppliers may also have limitations on the position of
dumping devices relative to the last stage blades. It is important that these requirements reach
the condenser supplier in the initial design stage.
To ensure the most reliable steam dump system it is incumbent upon the plant designer to
provide the condenser designer the operating scenarios expected. Otherwise the condenser
designer cannot check the range of conditions and off design operational problems could arise.
The condenser is a robust carbon steel encasement of a comparatively delicate tube bundle. The
tubes are the essential heart of the condenser and are subject to failure from waterside corrosion,
mechanical damage, vibration fatigue and erosion. Erosion on the steam side is a direct result of
wet steam at high velocity. Tube walls can be eroded through in a manner of minutes and high
water densities can promote buffeting thus producing fatigue failures.
Impingement shielding can be installed in a condenser to mitigate the effects of wet steam but
cannot ensure that erosion damage will be eliminated. They are at best wearing parts that will
require replacement in time and may deflect the erosion to an adjoining area. Impingement
shields by their nature can interfere with the normal steam flow and produce extra pressure drop
between the condenser and the tube bundle affecting turbine output negatively. The only true
and correct solution is to maintain dry and safely distributed steam to the condenser at all modes
of steam dump operation.
With a standard bypass arrangement consisting of a steam dump valve, attemperator, either
within the valve or separate and a perforated dump tube the sources of water in steam dump lines
have been attributed to:
Draining the steam lines that are prior to the dump valve through the dump valve and not
through a drain pot in the steam dump line.
Turning on the desuperheating water flow prior to establishing steam flow.
Locating attemporation devices too close to the condenser.
No drain provisions in the dump tube in the condenser.
Undrained loops in the steam dump piping to and in the condenser.
Desuperheating systems that do not provide monitoring of downstream conditions.
Draining steam lines through the dump valve is particularly dangerous from an erosion
standpoint on start up when steam flows are low and the temperatures are relatively low. At the
low pressures existing within the condenser there is virtually always sufficient pressure within
the steam dump tube to produce sonic velocity across the orifices. The water exiting the steam

1-32
dump device is easily accelerated to damaging velocities. Installing drain pots on the main steam
line prior to the dump valve will eliminate this problem.
Establishing desuperheating water flow prior to establishing steam flow can be devastating to the
condenser. If the desuperheating water flow is low only minor erosion may occur. If the
desuperheating water flow overwhelms the drain capacity of the dump tube major damage in the
condenser can occur as the excess water is expelled by the steam flow at near sonic velocities
first as solid streams then as droplets. Under certain conditions physical damage to the steam
dump tubes is also possible. Interlocking the desuperheating water valve with the steam dump
valve so that the desuperheating valve cannot open until the steam dump valve has opened can
eliminate this potential problem. It would also be advisable to program in a small time delay of
up to 3 - 5 seconds to ensure that the steam dump flow has been established prior to
desuperheating water flow.
The trend has been to move attemporation devices and/or valves closer to the condenser to save
plant costs. Attemportation is a time related function that is dependent upon droplet size and the
amount of remaining superheat. Thus complete evaporation and hence dry steam do not occur
until some distance down the pipe from the device. The general rule is that 90% evaporation
occurs at the distance calculated from 0.1 times the velocity in the pipe and 100% evaporation is
obtained at 0.3 times the velocity in the pipe. The actual distance to be used in design is
dependent upon the type of attemporation device and the actual operating conditions.
If complete mixing has not occurred by the time the condenser connection is reached erosion of
the dump tube holes and possibly condenser parts and/or tubes is possible. A guarantee should be
obtained from the attemporation device manufacturer, which may be included in the steam dump
valve, for the longest distance for complete mixing under all conditions of operation. Then if the
attemporation device is located at least this distance from the condenser connection, to ensure
100% evaporation, this potential problem can be eliminated.
Unheated parts of any system handling steam are subject to continuous condensation. Thus
water build up is insidious and must be foremost in the systems designers mind. Water
retention in the dump piping and/or the dump tube both outside and within the condenser is
easily rectified by ensuring the subject is part of the agenda for every design review and that low
points are designed out and adequate drains are designed in.
The trend today in desuperheating control systems is to install an enthalpy based digital control
system (DCS). The traditional method that has been proven successful and reliable over the
years was a temperature feedback system. The temperature feedback system by nature is based
on the temperature conditions down stream of the desuperheating device and the condensate
valve is adjusted to maintain the set temperature. Once set for the design temperature that
provides the appropriate superheat (i.e. typically 25F to 75F) at full load conditions, at all
lower load conditions the steam will always be dry. If there is an upset in the operation of the
desuperheating system it can be detected immediately since the downstream temperature is being
monitored.
One type of enthalpy based DCS takes as its inputs condensate temperature, condensate flow,
dump steam flow and dump steam enthalpy. Using these inputs and an algorithm the DCS

1-33
calculates the amount of condensate flow required to desuperheat the incoming dump steam flow
and then adjusts the condensate control valve until the condensate flow readings match.
The dump steam enthalpy is typically provided by a separate DCS that takes input from
temperature and pressure sensors in the main steam dump piping. This reading should be
relatively accurate. The same holds true for the condensate temperature, which again comes
from a direct temperature reading in the condensate line. The condensate flow is typically
monitored using an orifice type flow meter that should have a high degree of accuracy.
The only input value that is usually not directly known is the dump steam flow. This is typically
obtained from the output of the steam dump valve position indicator and the valve flow
coefficient (Cv) curve the valve manufacturer supplies for the valve. Thus the steam dump flow
input is subject to the accuracy of the valve Cv curve and the setting of the valve positioner.
Without a fairly accurate value for the actual steam flow the calculations this type of enthalpy
based control system could provide output conditions with excessive superheat or excessive
moisture.
With the use of this type of enthalpy based control system the conditions downstream in the
steam dump tube are never really known only assumed. This inability to know the actual
conditions in the dump tube means that the plant operators never know if the system is operating
properly or if there is an upset. Thus these types of systems should have at least a temperature
sensor downstream to monitor the conditions and assure proper operation. Which begs the
question Why change from the proven temperature feedback system? All steam dump control
systems regardless of type should monitor the down stream conditions to preclude damage to the
condenser.
Regardless of the type of desuperheating control system used, dry steam at start up and low
flows is directly related to the ability of the desuperheating condensate control valve to
effectively control the flow at the low end of the range. Particular attention to valve trim is
critical to allowing proper full range operation and preventing excessive temperature or moisture
in the steam dump lines.
Probably the largest single factor to reduce the reliability of condensers with steam dump has
been the continual pressure of the plant designers to lower the overall plant height to benefit
from the attendant cost reductions. This results in reducing the steam dome height on down
exhaust condensers and the volume necessary to dissipate the thermal and kinetic energy to safe
levels. Down exhaust condensers are where the majority of problems have occurred. There have
also been several technical papers written over the years on the subject of providing the
appropriate steam dome height. The Electric Power Research Institute has even promulgated
guidelines in this area. Yet economics may overshadow reliability.
The appropriate steam dome height for properly handling the steam dump is combination of the
physical number and sizes of steam dump lines, the actual steam dump flows and the placement
of the steam dump tubes within the steam dome. Dump tubes must be far enough away from the
tubes and condenser internal structures to prevent damage from the sonic core and expansion
cone of the steam jets. There must be sufficient distance and volume to prevent localized high
velocities and provide time for the condenser desuperheating sprays to reduce the temperature.

1-34
For the reasons stated above the location of the steam dump tube within the steam dome is one of
the most important factors in providing a reliable design. To assure proper operation the plant
designers must rout steam dump piping to the condenser entrance points as requested by the
condenser designer with the reliability of the condenser in mind not, as is often the case, for least
cost and/or plant design convenience.
In the preliminary plant design stages it is difficult to determine the appropriate steam dome
height. The following guideline will allow adequate space for most cases. Figure A depicts a
typical down exhaust condenser with two dump tubes in the steam dome.
To prevent erosion of the side walls no dump tubes should be located closer than 6 feet to the
edge of the steam dome. This is depicted as Min. Dist. on Figure 1.
The height above the top of the tube sheet to the first dump tube horizontal centerline is
determined by multiplying the Max. Dist. by .4 and adding 24. The Max. Dist. is the
distance from the dump tube vertical centerline to the farthest condenser side wall. If the
dump tube is on the condenser centerline this dimension is one half the shell width. If not on
the condenser centerline this dimension is always greater than one half the shell width.

1-35

24"
D2
D1
Max. Dist.
From dump
tube to
farthest
condenser
side wall
Min. Dist.
From dump
tube to cond-
enser side
wall 6' - 0"
24" + 1/2 D2
Minimum
18" + 1/2 D1 + 1/2 D2
Minimum
.4 x Max. Dist.
FIGURE 1

1-36
If there is more than one dump tube and they are arranged one above the other than the
minimum distance between their horizontal centerlines is one half the diameter of each plus
18.
If there is more than one dump tube and they are arranged side by side the distance between
the vertical centerlines is six(6) feet minimum.
The distance between the highest dump tube and the top of the steam dome is one half the
diameter of the highest dump tube plus 24.
The same guideline can be used for axial or side exhaust condenser arrangements by turning the
diagram 90 and referencing the Min. Dist. and Max. Dist. dimensions to the top and bottom of
the steam dome rather than to the condenser side walls.
The above safe practice guidelines are based on perforated dump tubes. Other types of steam
dump distribution systems could have different requirements for the necessary steam dome
volume to safely distribute the steam dump.
By applying these requirements for combining steam dump and condensers I am hopeful that
history will again repeat itself, this time in a positive manner and steam dump problems will
again become a part of the past.
References:
1. EPRI Symposium on State-of-the-Art Condenser Technology, June 1983 in total and in
particular Space Allotment for Surface Condensersby Bow, W.J. (Foster Wheeler)
included therein.
2. Heat Exchange Institute, Inc. Standards for Steam Surface CondensersNinth Edition, by
Heat Exchange Institute 1995.
2-1
2
SESSION 2: MATERIALS TECHNOLOGY
Thin-Wall Titanium Condenser Tubing: The Next Plateau
Dennis J. Schumerth
Valtimet, Inc.
The Performance of Superferritic Stainless Steels in High Chloride Waters
John C. Tverberg, P.E., and Edward R. Blessman, P.E.
Trent Tube
A Main Condenser Success Story
Jim Mitchell
Plastocor, Inc.

Session 2: Materials Technology
2-3
THIN-WALL TITANIUM CONDENSER TUBING
THE NEXT PLATEAU
Dennis J. Schumerth
Valtimet, Inc.
Tustin, CA
Abstract
Commercially pure (cp.) titanium and its alloys provide excellent resistance to general and
localized corrosion attack under most oxidizing, neutral and inhibited reducing conditions in
aqueous environments. Titanium is also notable for its outstanding resistance to chlorides and
other halides generally present in most process streams. In addition, titanium resists other
malicious phenomenon including steam and particle erosion, crevice corrosion, galvanic attack
and MIC. Given this general corrosion immunity, designers have increasingly applied thin-wall
condenser tubing in pursuit of cost savings and performance enhancement.
Typically, these thin-wall applications have, over the past several years, been limited to 25 BWG
or 0.020"/0.5mm walls or heavier. The "industry" has, out of necessity, moved to address the
special nuances of the 25 BWG including design, procurement, handling, fabrication and testing
parameters with increasing success. It would appear however, that designers, operators and
pundits alike require further education and refinement on the specific operational characteristics
when integrated into the powerplant environment.
Considerable work has been recently completed investigating even thinner wall titanium tubing.
This paper will address the essential data elements of this expanded research focusing
specifically on Grade 2 titanium in 27 BWG or 0.016"/0.4mm. Since a significant portfolio of
27 BWG installations is rapidly taking shape, it is prudent to examine key ingredients that would
warrant consideration of this "next-plateau" gauge material. Indeed, work is currently underway
exploring the technology required to fabricate condenser tubes as thin as 30 BWG /0.013"/0.3
mm.
In summary, the paper will present and summarize substantive evidence suitable for comparison
against previously acquired empirical data and prior art.
Background
Previous work, completed during the mid 90's and published in 1999
(1, 2, 6, 7)
,

suggested the
properties of thin-wall titanium could provide attractive options to the designer/user in terms of
habitat suitability, long-term reliability and performance and economic savings.
Empirical and actual testing was used at that time to compare the prevailing wisdom of heavier
wall tubing usage vs. the newly introduced thin-wall.

2-4
Actual testing involving fatigue properties, both from internal pressure and external excitation,
demonstrated highly admirable characteristics. Buckling strength or longitudinal compression
tests were evaluated employing Euler's equations and found to be well within acceptable limits
(1,6)
. Other key issues including mechanical expansion and resultant pullout strengths were
compared against heavier wall tubing. The results suggested a rolled and welded joint should be
employed when considering tube walls less than 22 BWG/0.028./0.7mm. This procedure is
highly recommended for both solid and clad tubesheets. Support plate spacing was evaluated
employing steam-loading calculations applied to design base, bundle out of service and turbine
bypass conditions.
Data is now available which can directly compare the previous papers' investigations and
findings
(l)
against more mature, practical and demonstrable findings. In addition, the installation
experience at the time of the paper's presentation, which was limited to several installations in
Japan and Europe plus a host of desalination units, was considered inadequate.
Pull-out Loads
One of the most important elements of this paper's investigation is a comparison and
confirmation of the suitability, or lack thereof, of tube-to-tubesheet loads in a commercially
repeatable environment using the 27 BWG tubing. Previous testing, completed in 1999 (Figure
1), confirmed that pullout loads, employing a mechanically expanded joint alone, were not
sufficient to support the necessary safety factors required by the designer. However, when tube
welding was added to supplement the rolled-only joint, acceptable pullout loads resulted. The
tube parameters identified in the below Table 1 were used to develop the Figure 1 results.

2-5

TABLE 1

Tube Material Titanium Gr. 2 B-338
Tubesheet Material Titanium Gr. 2 B-265
Tube Size 1"/25.4mm
Drill-Out Hole 1.1"/28mm
Tubesheet Hole Plain - No Serrations
Joint Configuration Mech Exp or Exp & Weld

Additional pullout testing was recently completed which, when evaluated, could be directly
compared against results reported in the previous work. In this case, this new testing was
completed using the three, expansion/weld processes identified in Table 2 and tube/tubesheet
parameters noted in Table 3.

Figure 1
Tube Pull-Out Loads Tests - 1999
0
1000
2000
3000
4000
5000
6000
7000
30 BWG 27 BWG 25 BWG 22 BWG
P
u
l
l
-
o
u
t

L
o
a
d
s

(
l
b
s
.
)
Rolled Only Roll & Weld

2-6
TABLE 2

TEST NO PROCESS
1 MECHANICAL EXPANSION
2 MECH EXP + ID GROOVE ASSISTANCE
3 MECH EXP + TIG WLD. (NO ID GROOVE)

TABLE 3

Tube Material Titanium Gr. 2 B-338
Tubesheet Material Titanium Gr. 2 B-265
Tube Size - OD & BWG .866"/22mm x .016'/0.4mm
Drill-Out Hole .875"/22.225mm
Tubesheet Hole See Table 2
Joint Configuration See Table 2

Figure 2
Tube Pull-Out Load Test - 2001
6
8
10
12
14
16
18
0 500 1000 1500 2000 2500 3000 3500 4000
Pull-Out Loads (lbs.)
%

W
a
l
l

R
e
d
u
c
t
i
o
n
Mech Exp Exp + ID Groove Exp + Weld

2-7
It is clear, upon examination of the new data scatter in Figure 2
(3)
, that pullout loads are
comparable to the results demonstrated in the previous 1999 work. This data similarity is not
surprising and suggests practical and achievable results in a manufacturing environment.
Tube-to-Tubesheet Expansion
Five-roll, mechanical expansion of the 27 BWG/0.016"/0.4mm thin wall titanium tube should
range from 7 to 12% wall reduction. The following classic formula should be used when
calculating the per-cent wall reduction.

% Wall Reduction = D - (DE -2T)
2T

where D = Hole Diameter
DE = Inner Tube Dia. after Expansion
T = Tube Wall

1. Nominal tube expansion should not exceed the tube OD by 3%.
2. Five (5) roll expanders are recommended where the OD/T ratio is greater than 25
(8)
.

where OD = Nominal tube OD
T = Wall thickness

Expansion beyond the upper limit may actually reduce pullout loads and potentially, induce tube
cracking or incipient failure of the mechanical joint. In fact, historical results employing heavier
wall tubing suggest the percent expansion need not be as high as the thin-wall counterparts - less
than 10%.
Tubesheet & Support Plates
Particular care must be paid to the fit and finish of the drilled tubesheet plate holes. When
considering thin wall titanium, this author suggests tubesheets be drilled to TEMA Close Fit
Tolerance. This operation prevents excessive tube-to-tubehole clearance, which may induce
undesirable tube deformation during the rolling operation.
ID groove assistance employing multiple serrations or concentric rings clearly enhances pullout
strength. Serrations to a depth of 0.004"/.1mm or less are typical. Others prefer ID surface
enhancement or controlled roughness (< 50 microns R
z
) to achieve the desired results. What is
ultimately used to enhance pullout and sealing should be left to the discretion of the designer
based on actual testing and results.
Support plates should be drilled to HEI standards and deburred or chamfered on both sides of the
plate. Indeed, wire brush operations may not suffice in terms of complete removal of chip
material left behind after drill bit exit. This "both sides" operation is strongly suggested to

2-8
eliminate or at least minimize any scratches or "exit wounds" that could be imparted to the tube
OD. It is presumptuous to assume that support plate orientation during cage assembly will
segregate all one-sided drilled plates in the direction of tube entry. Hoping the tube bullet will
break any burr on the un-chamfered side is a bold assumption indeed and could result in
unanticipated problems associated with subsequent base line E-C or other post-assembly testing.
Tube-to-Tubesheet Welding
Tube-to-tubesheet welding, employing a TIG process with shielding gas, is strongly
recommended when using thin-wall tubing (under 22 BWG/0.028"/0.7mm) and solid titanium
tubesheets. It is imperative when employed in a clad or bi-metallic tubesheet arrangement. Be
advised that titanium can only be welded to titanium - no other commercially available material
is metallurigically compatible.
Increasing the tube protrusion beyond the face of the tubesheet and reduced weld-bead heat input
to the tube-tubesheet interface may prove successful in enhancing the nugget configuration and
quality. However, these operations may influence weld speed resulting in increased tube-to-
tubesheet weld time. Understand the thinner the tube wall, the more the tubesheet will become an
ever-larger heat sink. In addition, the more the tube wall is reduced, the higher degree of
difficulty in repairing the weld.
Vibration
Operational nuances, not normally encountered with more traditional designs, may come into
play more often when evaluating the use of thin wall titanium. These phenomenon are deserving
of special consideration by the designers and would include, but not limited to, the following.

2-9

Peripheral Tubes
Heavier walled tubes, typically reserved as "optional" or for the first several rows only, may require a deeper
bundle penetration.
Larger OD Tubes
The use of 1"/25.4mm or larger tubes may reduce the vibration potential.
Flow Induced Vibration
Steam flow may penetrate deeper into the bundle requiring close examination of flow-induced excitation
parameters.
Bundle Flow Areas
Entrance and exit areas, steam lanes, etc. in and around the bundle may have to be increased requiring
additional surface area.
Support Plates
As many as one or two additional support plates may be required.
Steam bypass & bundle-out-of-service
Anomolies associated with steam bypass conditions and off-design operation will require careful "what if"
scrutiny.

In terms of the support plate spacing, earlier experimental work and empirical data
(1)
, concluded
support plate spacing should be reduced as a ratio of tube wall vs. HEI allowable deflection. In
the case of 27 BWG tubing, a ratio of .76 determined the minimum spacing to be 29". It would

2-10
now appear that support plate spacing for 27 BWG tubing should fall into a range closer to the
24".

Handling and Logistics
To date, there is no evidence that suggests any greater degree of difficulty in handling the 27
BWG tubing vs. the heavier walled cousins. However, it is recommended that tube boxes be
placed in close proximity and elevation to the bundle entry area to prevent the possibility of
kinking or bending which can result from excessive tube movement. Care must be exercised to
prevent any denting during the fabrication process, as this stress riser could eventually become
an incipient point of high cycle fatigue failure.
Protection of the exposed tubing in high velocity and peripheral areas of the bundle must also be
addressed. Dummy tubes, impingement rods and other measures must offer complete protection
for the tubing from mechanical impact damage.
Tube Manufacturing
Tubing produced by the supply mills to the standards of ASTM B-338 must adhere to the highest
levels of quality but result in acceptable levels of mill productivity and yield. All this must take
place before the condenser fabricators are convinced that the material is suitable for use in a
surface condenser. Grade 2, 27 BWG titanium tubing has been successfully produced for a
number of years. Yet in terms of productivity, yields have been less than desirable. However,
over the past several years, full production runs of the strip material supplied from numerous
sources have proved highly successful setting the standard for commercial use.

Figure 3

2-11
Of particular significance are the low E-C (Electromagnetic or more commonly, Eddy Current
Testing) & UT (Ultrasonic Testing) rejection rates for the final product. Coupled with precise
control of the diameter, nugget geometry and weld-undercut discipline, users can be assured that
the product will perform in a manner meeting the requirements of ASTM B-338-01a.
Cleaning
In support of actual, in-service conditions, where fouling will most certainly take place, Conco
Systems of Verona Pennsylvania performed tests on sample titanium tubes manufactured to
ASTM B-338 Gr. 2.
Three tests were performed on the sample material (0.9450"/24mm outside diameter (OD) x 27
BWG/0.016"AVW/0.4 mm wall thickness).

Visual Inspection
E-C Examination
Hydrostatic Test
The tests were conducted using various Conco tube cleaning tools including type(s) C3S, C4S
and C3X.
A visual inspection was performed comparing an as-received sample section vs. a sample section
that had various tube cleaning tools "shot" or driven through. Each section was examined after
the test and compared against each other. No noticeable visual or dimensional differences were
discernable up to a 10X magnification.
An as-received sample was again examined using a Zetec MIZ-27 Analyzer. Only minor drift
signal anomalies were recorded probably associated with the tube weld. The signals were
considered insignificant. The same sample was again "shot" with the three (3) tube cleaning
tools identified above. The results noted no rejectable or signals of consequence. In fact, the
signals first recorded during the parent-tube test, actually reduced in magnitude after the samples
were cleaned.
Finally, a third sample was hydrostatically tested to a pressure of 2100 PSI. No failures occurred
noting only expected minor plastic OD deformation in the range of 0.015". In point of fact,
surface condenser tubes would never experience such high pressure. Even the use of individual
tube hydro testing, which operates in the range of 645 PSI, (ASME UG-27 max working
pressure calculation), falls well below either the 2100 PSI or a burst pressure of a 27 BWG
titanium tube (.016"/0.4mm) calculated at 3191 PSI/220 Bars.
Condenser tube fouling is real and a fact of life within an operating powerplant surface
condenser. It is apparent, after reviewing the test conditions and results, that this proven type of
cleaning system has no effect on the mechanical properties of the tube. If this axiom is true for
scraper cleaning systems, other, more benign cleaning methods including sponge balls, brushes,
water lances, etc. can be utilized with little or no apprehension.

2-12
Other Considerations
An HEI Ninth Edition Supplement, soon to be released, will address, among other issues, new
criteria to be applied for steam bypass conditions. Clearly, additional work is needed by this
august group beyond even this Supplement to address issues specific to combined cycle
applications. Notable and unfortunate as of this writing is the absence of any HEI Standards
information relating to metal resistance and/or correction factors for tubing less than 25 BWG.
Designers must therefore, address the heat transfer characteristics of 27 BWG tubing
independently. The thermal conductivity of titanium is readily available
(8)
(150 BTU/hr ft -
F/in) and the metal resistance can be easily developed from existing data in Figure 4.
(8)
should
alternate rating methods be employed. Designers will find that reducing wall thickness improves
the heat transfer characteristics of titanium when compared to other materials.
Reducing the wall thickness of the tubing reduces the weight of the tubing. It is noteworthy that
the weight of a 27 BWG titanium tube ranges between 55 & 60% (depending on diameter) when
compared to its 22 BWG counterpart.
These weight reductions and resultant uplift considerations require a more robust analysis of the
foundation loads, bolting, expansion joints and other condenser components that would be
impacted by the weight reduction. In addition, larger flow areas associated with thinner walled
tubes can influence circulating water flow and pump NPSH considerations.
Certain conditions may preclude the complete removal of existing condenser tubes or, plant
logistics make removal commercially impractical. In addition, a straight retube may prove more
costly, present dissimilar material constraints and modules do not provide the required payback.
Sleeving existing condenser tubes with ultra thin wall titanium is currently under test and
consideration. The thin wall is hydraulically expanded full length into the parent tube with the
tube ends mechanically anchored in each tubesheet. Should additional embellishment be
considered beyond just the sleeve replacement, a multi-part epoxy coating can be applied to add
corrosion resistance while enhancing the tube pullout load characteristics.
Installation History
The Japanese Titanium Society reported in earlier work,
(3,6,7)
successful installations of 27 BWG
Gr. 2 titanium in both desalination and powerplant applications. These installations have been in
service for over 15 years with no documented problems. Table 4 identifies more recent vintage
installations primarily in France and in the UK.
It is noteworthy however, that condenser manufactures in the US are now evaluating the use of
this gauge material

2-13
TABLE 4
27 BWG/.016"/0.4mm INSTALLATION LIST
(Partial)

YEAR OD KM DESTINATION

1981 17 5 FRANCE
1999
1999
1999
1999
2001
2001
2001
2001
2002

12.7
16
22
24
20
24
21
24
21
6
12
926
253
7
144
288
294
137
FRANCE
FRANCE
SPAIN
GB
FRANCE
MALAYSIA
VIETNAM
GB
PORTUGAL

Conclusion
Given continuing, industry-wide pressure to reduce costs, improve performance and significantly
contribute to improved online availability and capability, a transition to less costly, high
performance materials is inevitable. In the case of surface condensers employing 27 BWG or
0.016"/0.4mm tubes, the opportunity exists to examine, evaluate and realize the potential merits
of reduced or thin wall tubing.
Notable in the work just presented, and a key element in this paper's investigation, is a clear
confirmation of acceptable pull out loads. Welded-only tube joints, and/or welded joints
employing I.D. enhancement, clearly demonstrate the repeatable practicality of achieving an
acceptable tube joint within a shop manufacturing environment.
Mechanical expansion techniques, tubesheet and support plate fit & finish, tube-to-tubesheet
welding and other processes appear to be successfully practiced in today's environment. They are
clearly, not future technical folly but proven technology that exists today.
Efforts to address the nuances of combined cycle steam exhaust and bypass conditions will
require additional work by professional societies and the manufacturers alike. Current
techniques are frankly, not adequate to predict certain dangerous operating conditions.
Tube mill production runs of strip material supplied from numerous sources have produced
successful yield and production goals setting the quality standard for commercial use. Handling,
at both the supplier and fabricator levels, does not appear to present logistical issues nor invoke
any specialized or precautionary steps.
An analysis of cleaning techniques has confirmed the future suitability of both the tube and the
cleaning systems.

2-14
Recall the object of this paper, which was to evaluate additional, recently acquired data to
compare, validate and expand on prior work. This author believes that practical confirmation of
this work has indeed, been documented and achieved. Confirming the validity of prior art while
expanding directly into newly developed technology and practical issues provides merit for
future examination of thin wall titanium condenser tubing.

TUBE
O.D
IN/mm

BWG

WALL
THICKNESS
IN/mm

FACTOR
(FM)

Metal
Factor
(RM)

1/25.4
1/25.4
1/25.4
1/25.4

0.875/22.225
0.875/22.225
0.875/22.225
0.875/22.225

0.75/19.05
0.75/19.05
0.75/19.05
0.75/19.05

25
26
27
30

25
26
27
30

25
26
27
30
0.020/0.508
0.0180.457
0.016/0.406
0.013/0.330

0.020/0.508
0.018/0.457
0.016/0.406
0.013/0.330

0.020/0.508
0.018/0.457
0.016/0.406
0.013/0.330
0.95
0.96
0.97
0.99

0.95
0.96
0.97
0.99

0.95
0.96
0.97
0.99

1.3607
1.2222
1.0853
0.9867

1.3576
1.2232
1.0878
0.8787

1.3702
1.2293
1.0908
0.881

Notes:
Supplement to TIMET Tube Book Table 9 Metal Correction Factor (Fm)
Based on 0.875" O.D. @ 7 ft/sec C@ 70
0
F
Material = Gr. 2 Titanium
RM (Metal Resistance ) = 1 x 10
-4

BTU/hr ft
2

0
F
FM (Factor) Titanium Gauge Correction Factor
DMc 3/31/00

Figure 4
TITANIUM METAL RESISTANCE METAL CORRECTION FACTORS
SUPPLEMENT TO TIMET TUBE BOOK TABLE 10 - 1996 REV

2-15
References
1. Schumerth, - "Thin-Wall Titanium Condenser Tubing Explore the Opportunities." PWR -
IJPGC ASME 2000
2. Schumerth, McCue - "Titanium Surface Condensers & Heat Exchangers - A Practical Guide
and Application." PWR-Vol 33 IJPGC Volume 2 ASME 1998
3. HEI - Heat Exchange Institute
4. HEI Standards - Ninth Edition
5. ASME Boiler & Pressure Vessel Code Section UG-27
6. Japan Titanium Society - 1984 - Multiple Authors - et al. "Get More Advantages By
Applying Titanium Tubing Not Only For Power Plants But Also For Desalination Plants."
7. Japan Titanium Society - Multiple Authors - et al. Thinner Wall Welded Titanium Tubes For
Seawater Desalination Plant.
8. CONCO Systems - January, 2002 - Thin Wall Titanium Tube Testing
9. TIMET Codeweld Tube Book - 1984 (plus subsequent revisions)

2-17
The Performance of Superferritic Stainless Steels in High Chloride Waters
John C. Tverberg, P.E.
and
Edward R. Blessman, P.E.
Trent Tube
East Troy, Wisconsin

Abstract
Superferritic stainless steels were developed 25 years ago as a low cost alternative to titanium
and the high molybdenum austenitic nickel alloys for use in brackish and seawater. These alloys
have a ferritic structure with very high chromium and molybdenum content. The result is a
material with better corrosion resistance than the austenitic or duplex stainless steels, and
approximately the same as unalloyed titanium, thus they are applicable in almost all water
conditions. Their modulus of elasticity is the highest for the common engineering alloys. As a
result these alloys have seen wide scale use in condenser applications where vibration is a
problem and limited use in feedwater heaters and balance of plant heat exchangers. They have
better thermal conductivity than the superaustenitic and duplex stainless steels and only slightly
less than titanium. Their high hardness provides excellent sand and inlet end erosion resistance.
Introduction
Superferritic stainless steels are a rather recent development. For years engineers have had
interest in ferritic stainless steels because of their advantages over competing materials. Ferritic
stainless steels have excellent resistance to chloride pitting and crevice corrosion, they are
resistant to chloride stress corrosion cracking and they have excellent resistance to organic acids
and caustic environments. But they had one major problem: the normal carbon and nitrogen
caused low ductility and reduced toughness. It wasnt until the mid sixties that several new
technology developments allowed the production of stainless steel with low interstitial carbon
and nitrogen. These were electron beam melting (EBM), electroslag remelting (ESR), argon-
oxygen-decarburization (AOD) and vacuum-oxygen-decarburization (VOD).
Superferritics were developed in three phases. The first phase used high purity melting
techniques, vacuum induction furnaces and electron beam melting. The second phase involved
adding nickel to the alloys to improve the manufacturability; and the third phase used stabilizing
elements titanium and niobium (columbium) together with AOD refining to allow commercial
production.
The first superferritic, E-Brite 26-1, was introduced in 1970 by C. D. Schwartz, I. A. Franson
and R. J. Hodges of Allied Vacuum Metals
1
. It is based on the composition of Type 446 ferritic
stainless steel which contains 23 27% Cr. E-Brite 26-1 has an interstitial C+N content of 200
ppm attained only by a combination of vacuum induction melting followed by EBM or ESR.

2-18
This results in an improvement in the ductile to brittle transition from +120 C (250 F), for
Type 446 stainless to -60 C (-80 F) for E-Brite 26-1. This alloy possesses outstanding
corrosion resistance, especially to chlorides and strong caustic environments.
The success of E-Brite 26-1 prompted the development of a number of other alloys. The first
was 29Cr-4Mo by M. A. Streicher
2
at duPont who filed for patent several months after E-Brite
26-1 was introduced. In 1974 Climax Molybdenum introduced 18-2
3
and several months later
Deutsche Edelstahlwerke introduced 28Cr-2Mo. Despite the outstanding performance of these
alloys in chloride environments, they were not commercially practical because of the need for
double vacuum processing.
Work was undertaken at a number of steel mills to stabilize the alloys with titanium and/or
niobium and to use AOD refining to obtain the low carbon levels. R. Oppenheim and J.
Lennartz
4
of Deutsche Edelstahlwerke are thought to be the first with 28Cr-2Mo in 1974. In the
meantime Streicher developed 29Cr-4Mo-2Ni which was still produced by vacuum melting. This
alloy led to the development of Monit
, 26Cr-4Mo-4Ni, by Nyby-Uddeholm
5
in Sweden, SEA-
CURE
Stainless, 27-4-2, by K. E. Pinnow

6
of Crucible Research in the United States and 29-4C
by Allegheny Ludlum also in the United States. These alloys will be designated by their UNS
Numbers hereafter in this paper. Refer to Table I for the cross reference to the common
Trademarked names.
The Superferritics were introduced to the power market in 1979 for use in main steam
condensers. The first Superaustenitic condenser tubes, AL-6X, were installed in 1975 and the
first Superferritic condenser installation, 29-4C, was in 1974. Since then nearly 60,000,000 feet
of Superferritic condenser tubing has been installed. Figure 1 illustrates the cumulative
quantities of both superaustenitic and superferritic stainless steels that have been installed. Last
year the superferritic stainless steels surpassed the total installed footage of superaustenitic
stainless steel. Since 1998 82% of all high performance stainless steel tubing installations have
been superferritic, and the majority of that is UNS S44660.
This surge in condenser usage is a result of a combination of outstanding chloride induced
corrosion resistance, excellent sand erosion resistance and droplet impingement resistance,
excellent heat transfer properties, outstanding mechanical properties in conjunction with over 20
years of successful operation history, ready availability and attractive pricing.
Composition and Metallurgy
Superferritic stainless steels are characterized by high chromium, high molybdenum and are
stabilized with titanium and/or niobium. Compositions of the most common superferritics are
given in Table II.
The addition of nickel is an important development. Nickel lowers the ductile-brittle transition
(DBT) temperature so the alloy is ductile over a wider temperature range. Nickel additions
together with low interstitial carbon and nitrogen allow the DBT to be reduced to -120 F (-84
C) for UNS S44660. Nickel also improves the corrosion resistance to reducing acids.

2-19
If the carbon and nitrogen are high in the iron-chromium alloys, then formation of an austenite
loop in the high chromium alloys may form. This means martensite forms during cooling from
heat treating temperatures. Martensite is hard, brittle and has limited ductility. By keeping the
C+N 0.06% martensite formation can be prevented and the alloys retain a single phase ferritic
structure. The grains are uniform and equiaxed with no grain boundary precipitates.
Chloride Corrosion Resistance
Stainless steels derive their corrosion resistance from a very thin, in the range of 10 to 30 atoms
thick, surface passive layer consisting of Cr
2
O
3
and Fe
2
O
3
in which the chromium to iron ratio is
greater than 1.0. As long as this passive layer remains intact stainless steel is resistant to
corrosion attack. Chloride ion is the major corrodant of stainless steel. Pitting and crevice
corrosion are the two most common corrosion mechanisms involving chloride ion, followed by
chloride stress corrosion cracking in the low nickel austenitic stainless steels. Low pH makes
this corrosion attack worse. For power station condenser applications, chlorides are the most
common corrodant.
Monnartz
7
discovered the synergistic effect between molybdenum and chromium in extending
the corrosion resistance of stainless steel to acid chlorides. The superferritic alloys require more
chromium and less molybdenum to accomplish this as compared to the superaustenitic stainless
steels. The most common method of testing the stainless steels is the use of ASTM G 48, which
uses ferric chloride and usually at pH less than one. The crevice corrosion test, Practice D, is the
more aggressive, especially at temperatures over 40 C (100 F). Since crevice corrosion is
temperature dependent, a convenient way to rank alloys is the critical crevice corrosion
temperature (CCCT), the temperature above which crevice corrosion takes place.
A convenient method of estimating relative corrosion resistance is to use the pitting resistance
equivalent number (PREN) developed by Rockel
8
and defined as
PREN = %Cr + 3.3(%Mo) + 16(%N).
Kovach and Redmond
9
combined the PREN with the CCCT and created a chart which allows a
comparison of the relative resistance of any composition of stainless steel to the temperature at
which crevice corrosion starts. Figure 2 is a modification of the original chart.
A ranking of the relative chloride resistance of selected stainless steels is presented in Table III.
In this table the alloys will be resistant to crevice corrosion at temperatures and chloride contents
below those stated. This table indicates that a superferritic with essentially the same PREN will
have a critical crevice corrosion temperature higher than an equivalent superaustenitic stainless
steel.
Pitting corrosion is not as severe as crevice corrosion and the critical pitting temperature is not as
clearly defined. In most cases starts approximately 40 C (100 F) above the critical crevice
corrosion temperature. Figure 3 illustrates the effect of pH, chloride content and alloy
composition on the pitting resistance of austenitic stainless steels with various molybdenum
contents. The superferritic stainless steels are included on this chart, but will lie between the 6%
and 9% Mo lines.

2-20
The other problem with stainless steel and chloride service is stress corrosion cracking. Again,
there is a threshold temperature, below which the alloy will not crack, above which it will. In
general, this threshold temperature increases with molybdenum content. For S30403 stainless
steel the threshold temperature is at room temperature, 20 C (70 F), S31603 stainless 50 C
(125 F), N08367 225 C (450 F), and S44660 225 + C (450 +F).
Microbiologically Induced Corrosion (MIC)
Superferritic stainless steels are resistant to MIC caused by sulfuric and sulfurous producing
bacteria. This is critical in some condenser locations. Of particular interest, especially to power
stations along the Ohio, Tennessee and southern Mississippi Rivers, is the resistance to
permanganate in the presence of chloride. This combination of chemicals is generated when
condensers containing colonies of manganese fixing bacteria, usually gallionella, are present and
the system is chlorinated to destroy the colonies. The chlorine or hypochlorous acid reacts with
the manganese dioxide deposits, the metabolic by-product from the bacteria, to produce
hydrochloric acid. A more complete dissertation on this reaction is given elsewhere
10
. The
superferritic stainless steels are resistant to the action of the hydrochloric acid.
Erosion and Steam Impingement Resistance
Both the superferritic and superaustenitic stainless steels have outstanding resistance to steam
side droplet impingement, cavitation, turbulence and high velocity flow. These alloys resist both
mechanical damage and the tendency for flow to accelerate corrosion. Their high mechanical
strength is the major factor in resisting mechanical damage due to flow. The durability of these
alloys is further enhanced by their ability to be significantly work hardened. Under the local
strain of impingement or flow the surface of these already strong materials will work harden to
even higher levels. Resistance to flow accelerated corrosion is largely determined by the nature
of the passive layer on the surface of the material. Alloys such as these high performance
stainless steels form thin hard passive films. This results in excellent resistance to accelerated
corrosion due to flow in these alloys. Table IV compares the different classes of alloys for these
parameters.
Mechanical and Physical Properties
Superferritic stainless steels are characterized by high strength, the highest modulus of elasticity
of any common heat exchanger material, good elongation, high hardness, good thermal
conductivity and excellent fatigue endurance. The coefficient of thermal expansion is close to
that of carbon steel, so the expansion and contraction of the tubes in the condenser will be nearly
the same as the carbon steel shell, resulting in minimal deflection of the tube sheets. A
comparison of the different alloy types is presented in Table V. The comfortable spread between
the yield strength and ultimate tensile strength combined with high ductility makes these alloys
very easy to work. The high strength not only imparts resistance to damage during operation, but
also reduces the risk of installation related damage.

2-21
Vibration Resistance
Vibration is a problem in all low pressure condensers. The maximum length between support
plates is determined by the equation
L = 9.5[(EI)/v
2
D]
1/4

where L is the unsupported length, E is the modulus of elasticity, I is the moment of inertia, is
the turbine exhaust density, v the average exhaust steam velocity at the condenser neck and D is
the tube diameter. The variation in tube spacing for different alloys becomes a ratio of the
modulus of elasticity times the moment of inertia (EI). The moment of inertia is a function of
the tube wall thickness. Thus, as the modulus of elasticity decreases the tube wall must increase
accordingly to maintain the same unsupported tube length.
Table VI compares a number of alloys with the normal wall thickness used in condenser
applications. This comparison was made on the basis of Admiralty Brass as the standard length.
When the relative span becomes longer, the tube is immune from vibration, as it becomes less,
the tube becomes susceptible to vibration. In the case of C70600 and R50400 the tubes would
need to be staked to prevent vibration. The superferritics would not need staking, nor would they
vibrate, assuming the C44300 tubes are stable.
Thermal Performance
Superferritics have excellent thermal performance in steam condensers compared to other
stainless steels and have similar performance as R50400. The heat transfer performance, based
on the HEI method of calculation, is essentially the same as that of the copper nickel alloys
because of their good thermal conductivity, thin wall thickness and ability to maintain
cleanliness. Superferritics also have the potential for improved heat transfer by increasing the
water velocity through the tubes. Since the water velocity typically is established to prevent
erosion of copper based alloys, it is possible to increase the velocity significantly to improve heat
transfer without the danger of inlet end erosion of the superferritic tubing.
Table VII compares the heat transfer for a number of different tube alloys. The significant point
here is the turbine back pressure and the fact that the superferritics can maintain a very low back
pressure.
Tubesheet Materials
The main criterion for tubesheet materials is its ability to resist crevice corrosion. For this reason
one of the 6% molybdenum alloys is the best choice for use in seawater. For less aggressive
environments S22205 is a good choice as is one of the 4% molybdenum alloys. Alloy S31603
should never be used because of the risk of dissimilar metal crevice corrosion and the attack of
both the tubesheet and tubes. Superferritic tubes can be installed into virtually any tubesheet
material. However, if the tubesheet is not galvanicly compatible or fully resistant to crevice
corrosion in the service environment, then the tubesheets must be coated.

2-22
Eddy Current Testing
Superferritic stainless steels can be eddy current tested in the condenser, however they require
full magnetic saturation. Therefore testing must incorporate an electromagnet or a permanent
magnet. In theory this material should be able to be tested using Flux Leakage methods. So far
no one has been able to demonstrate that they can achieve the necessary sensitivity with Flux
Leakage. Many testing service providers have successfully eddy current tested these alloys.
However, because of the added difficulties in testing and interpretation it is suggested that testing
vendors be required to demonstrate their proficiency.
Alloy Limitations
Superferritic stainless steels have these limitations:
1. They are sensitive to hydrogen embrittlement. If the condenser is operated with a cathodic
protection system, the voltage must never exceed 0.80 volts. If embrittlement does occur, the
reaction is easily reversible by heating to 100 F (40C) or allowing the tubes to sit in air for
24 hours.
2. They are sensitive to 885 F (475 C) embrittlement. This starts at about 600 F (315 C)
and is cumulative. However, they can be operated at 550 F (300 C) indefinitely without
danger of embrittlement.
3. Superferritic stainless steels are subject to embrittlement if they are operated at temperatures
below -40 F (-40 C). This better not be a problem with a steam condenser.
Summary
Superferritic stainless steels are rather new, lower cost addition to the heat transfer engineering
world, but in their brief service time they have made a definite impact. Their excellent corrosion
resistance, especially with respect to acid chlorides and MIC attack, have made them attractive
alternatives for steam condenser service. The modulus of elasticity, the highest of all common
engineering alloys, provides excellent vibration resistance. Their thermal expansion coefficient
is close to that of carbon steel so bending stress on the tubesheet is minimized. Most important,
their good thermal conductivity and ability to maintain cleanliness means they can maintain a
low turbine back pressure.
References
1. C.D.Schwartz, I.A.Franson, R.J.Hodges, Chemical Engineering, 77, April 20, 1970, Pages
164-167
2. M.A.Streicher, Corrosion, 30, (3), 1974, 77-91
3. M.Semchyshen, A.P.Bond & H.J.Dundas, Toward Improved Ductility and Toughness,
symposium sponsored by Climax Molybdenum, Kyoto, Japan, Oct. 25-26, 1971, 239-255

2-23
4. R. Oppenheim, G. Lennartz, H.Laddach, TEW-Techn. Ber., 2 (1), 1976, 3-13
5. N. Pessall & J.I. Nurminen, Development of Ferritic Stainless Steels for Use in Desalination
Plants, Corrosion, 30, (11), 1974, 381
6. K.E. Pinnow, Progress in the Development of High Chromium Ferritic Stainless Steels
Produced by AOD Refining, Stainless Steel 77, London, England, September 1977
7. P. Monnartz, Metallurgia, 8, 1911, 161-176 and 193 - 201
8. M. Rockel, Use of Highly Alloyed Stainless Steels and Nickel Alloys in The Chemical
Industry, ACHEMA Conf., Frankfurt, Germany, 1978
9. C.W.Kovach and J.D. Redmond, Corrrelations Between The Critical Device Temperature,
PRE-Number, and Long Term Crevice Corrosion Data for Stainless Steels, Corrosion 95,
Nat. Assoc. of Corrosion Engineers, Paper 267
10. John Tverberg, Kenneth Pinnow and Lawrence Redmerski, The Role of Manganese Fixing
Bacteria on the Corrosion of Stainless Steel, Paper 151, National Association of Corrosion
Engineers, Corrosion 90, April 23-27, 1990

Table 1
Superferritic Stainless Steels by Name and UNS Number
Common Alloy Name UNS Number
Type 446 Stainless S44600
E-Brite 26-1 S44627
18-2 Stainless S44400
29-4 S44700
29-4-2 S44800
Monit S44635
AL 29-4C S44735
SEA-CURE Stainless S44660

Table 2
ASTM Chemical Composition of Superferritic Alloys
UNS
Number
C Cr Mo Ni
Max
Si
Max
N
Max
Ti/Nb
Max
S44600 0.012 23.0-30.0 -- -- 0.75 0.10-0.25 --
S44627 0.010 25-27.5 0.75-1.50 0.5 0.40 0.015 0.05-0.20 Nb
S44400 0.025 17.5-19.5 1.75-2.50 1.00 1.00 0.035 0.80
S44700 0.010 28.0-30.0 3.5-4.2 0.15 0.20 0.020 --
S44800 0.010 28.0-30.0 3.5-4.2 2.0-2.5 0.20 0.020 --
S44635 0.025 24.5-26.0 3.5-4.5 3.5-4.5 0.75 0.035 0.80
S44735 0.030 28.0-30.0 3.60-4.20 1.00 1.00 0.045 1.00
S44660 0.030 25.0-28.0 3.0-4.0 1.0-3.50 1.00 0.040 1.00

2-24
Table 3
Maximum Chloride Resistance at Temperature for Various Stainless Steels
Alloy UNS Number PREN CCCT Max Chloride
Type 439 S43035 18 < -2 C (28 F) 100
Type 304L S30403 19 < -2 C (28 F) 100
Type 316L S31603 24 < -2 C (28 F) 500
Type 317L S31703 28 2 C (36 F) 1000
Type 317LM S31725 32 16 C (61 F) 2500
Type 317LMN S31726 33 20 C (68 F) 5000
Type 2205 S32205 34 24 C (75 F) 7500
AL-6XN
N08367 45 45C (113 F) 18000

SEA-CURE
S44660 46 52 C (125 F) 20000+

Table 4
Relative Resistance to Flow, Erosion and Impingement
Alloy Type Max. Flow in Sea Water,
ft/sec
Cavitation and
Turbulence
Steam Side Droplet
Impingement
Copper Alloys
7 Poor Poor
Copper-Nickel 11 Marginal Marginal
300 Series Stainless >125 Good Good
Superferritics >150 Very Good Very Good
Titanium 120 Good Acceptable

Table 5
Comparative Properties of Alloys in Sea Water
Property 90-10 Cu-Ni
C70600
70-30 Cu-Ni
C71500
AL-6XN

N08367
SEA-CURE

S44660
Ti Grade 2
R50400
Ult. St. ksi
40 52 100 85 50
Yld. St. , ksi 15 18 45 65 40
Elong. % 25 30 20 20
Hard. HB 20 22 100 104 92
Mod. Of Elas. Psi 18.0x10
6
22.0x10
6
28.2x10
6
31.5x10
6
14.9x10
6

Density,lb/ci 0.32 0.323 0.29 0.278 0.16
Thermal Exp. 9.5 8.2 8.7 5.38 5.20
Thermal Cond. 26 17 7.5 10.3 12.5
Fatigue Strength, ksi 20 25 33 35 16
Corrosion in Still
Seawater @ 65 mpy
2.1 0.67 <0.001 <0.001 <0.001
Max Flow in Seawater
fps
8 10 >150 >125 120

2-25
Table 6
Relative Spans For Different Alloys and Wall Thickness
Alloy Nominal Wall Thickness Relative Span
C44300 Admiralty Brass 0.049 in. (1.24 m) 1,000
C70600 90-10 Cu-Ni 0.035 in. (0.89 m) 985
R50400 Ti Grade 2 0.025 in. (0.64 mm) 515
N08367 AL-6XN 0.028 in (0.71 mm) 1,035
S44660 SEA-CURE 0.028 in (0.71 mm) 1,160

Table 7
Thermal Performance of Various Alloys in Condenser Service
Alloy
Gage U Cf Ps HL
C44300 Admiralty Brass 18 595 0.85 1.67 16.02
C70600 90-10 Cu-Ni 20 536 0.85 1.73 13.13
N08367 AL-6XN 22 531 0.95 1.73 11.92
R50400 Ti Grade 2 22 568 0.95 1.70 11.92
S44660 SEA-CURE 22 550 0.95 1.71 11.92
U = Overall Heat Transfer
Cf = Cleanliness Factor
Ps = Saturation Back Pressure
HL = Head Loss

2-26
Figure 1
Installation History of High Performance Stainless Steel Condenser Tubing
0
20,000,000
40,000,000
60,000,000
80,000,000
100,000,000
120,000,000
1972 1976 1980 1984 1988 1992 1996 2000
C
u
m
u
l
a
t
i
v
e

F
o
o
t
a
g
e
SuperFerritic SuperAustenitic Total

2-27

Figure 2
Critical Crevice Corrosion Temperature as a Function of PRE-Number showing the
variation according to alloy classification

2-28

Figure 3
Pitting corrosion as a function of chloride content, pH and molybdenum content of
austenitic chromium alloys. The superferritics will lie near the 9% Mo line. Temperature
range 150 - 180 F (65 - 80). Pitting not a problem below the line, but may be severe
above the line.

2-29
A MAIN CONDENSER SUCCESS STORY
Jim Mitchell, Plastocor, Inc.
Bernard Sloan, Dominion Virginia Power

Surry Power Station is a two-unit, three-loop pressurized Westinghouse reactor located on the
James River in southeast Virginia. Both 850 Mw units are owned and operated by Dominion
Virginia Power.
The station has implemented an extensive Secondary Plant Piping Replacement Program due to
flow-accelerated corrosion (FAC). Along with the pipe replacement program came a request
from Dominion Upper Management to resolve FAC issues.
The objectives of the replacement program include:
Increase plant personnel safety
Reduce material costs associated with replacement piping
Reduce manpower costs projected over the next five years to complete the piping
replacement.
FAC was also affecting the stations secondary side chemistry values for iron transport. Iron
removed from the secondary plant piping typically ended up in the steam generators. The plant
was well aware that the iron values were well in excess of industry median values of 3.3 ppb.
Once the iron reaches the steam generators, it has to be removed via sludge lancing to prevent
crevice corrosion in the tube-to-tubesheet joint interface.
In order to reduce iron transport, the pH of the Secondary Plant water chemistry had to be
increased. This increase in pH would induce a scaling/protective coating on the pipe ID, helping
to protect the piping from FAC. An amine additive was used to raise the pH.
Limiting the ability to raise pH, was the Condensate Polishing System used to purify and control
feedwater makeup. If the amine were increased solely to induce a more alkaline state, it would
result in an increase in condensate polisher vessel run times (resin regeneration frequency). This
increases: manpower costs, equipment wear (i.e. chemical pumps, valves, controls, etc.) and
resin requirements.
The Secondary Plant utilizes demineralized water for condensate makeup. Condensate
contamination generally occurs via sodium and chloride intrusion from tube-leaks and tube-to-
tubesheet joint leakage in the Main Steam Surface Condenser. Thus, the Condensate Polishers
primary function is to protect the Steam Generators from impurities coming from the main
condenser.
The Surry Main Condensers were originally installed with 90-10 CuNi tubes. Because of issues
relating to copper carryover and tube reliability, Unit #2 was retubed in October 1979 and Unit

2-30
#1 was retubed in March 1981 with titanium tubes. The titanium tubing was roller expanded into
the existing aluminum bronze tubesheets. The waterboxes were rubber lined.
The James River supplies brackish cooling and service water to the plant, with sodium levels as
high as 5,000 ppm. These high levels presented a serious corrosion concern for the plant.
Combining titanium with aluminum bronze in brackish water created the potential for galvanic
corrosion. The plant investigated various means for protecting the tubesheets (anode) from the
more noble titanium tubes (cathode). Impressed current cathodic protection and protective
coatings were identified as possible solutions.
The plant decided to use protective coatings. The first project was to epoxy-coat four (4) inlet
tubesheets on Unit 1 to prevent erosion-corrosion of the tubesheets and galvanic corrosion of the
tube-to-tubesheet joints. The four (4) outlet tubesheets were left un-coated. The coating had the
abrasion resistance and chemical properties sufficient to withstand the cooling water
environment. The condenser tubesheet coating project was designed to have a single supplier
responsible for materials and application. It was anticipated this would best insure conformance
to the coating application specification.
After significant run-time, in-leakage of the cooling water to the steam-side of the condenser was
experienced. This was due to outside diameter steam impingement on the peripheral condenser
tubes. These tubes were permanently plugged. In-leakage was also experienced due to tube-to-
tubesheet joint failure. This was attributed to failures of the tubesheet coating. In subsequent
outages, repairs to the tubesheet coating were made. The repairs became more extensive over
time. Figure 1 shows the spot failures of the coating.

Figure 1

2-31
A tubesheet coating that is failing creates a potentially more adverse corrosion environment.
With small areas of the tubesheet exposed or with sections of the coating having lost adhesion,
the cooling water came into direct contact with the tubesheet. This created a more concentrated
corrosion environment. An example of this selective corrosion is shown in Figure 2 where up to
of aluminum bronze tubesheet metal was lost. (The Figure 2 photo is NOT taken from the
Surry condenser but from another nuclear unit with the same metallurgy) It illustrates the
potential corrosion damage that can occur from a failing tubesheet coating. Even if impressed
current cathodic protection had been used as a back up to the coating for corrosion protection,
cathodic shielding of the tubesheet by the delaminating coating could impede the effectiveness of
the cathodic protection system.

Figure 2
Surrys decision was to remove the coating and replace it with an ultra high-reliabilty tubesheet
epoxy cladding that had gained a near flawless track record. In March 1992, four inlet
tubesheets on Unit #1 were re-coated with the Plastocor Tubesheet Cladding System. The 100%
solids epoxy materials were applied in a thick film (200 mils).
This coating system was applied thickly in an over-engineered manor to convey the required
reliability to prevent spot failures. The system is installed on a turnkey basis with one vendor
responsible for all materials, equipment, supervision, and labor. This approach assures the
quality and timeliness of the installation. A five-year labor and material zero defectwarranty
was also provided. References to substantiate successful completion of project performance
came from Virginia Power Fossil Installations and nuclear condenser installations outside of
Virginia Power.
Figure 3 shows the thick-film nature of the cladding. Figure 4 shows a schematic of the thick-
film claddings ability to rebuild the tubesheet by replacing lost metal with epoxy.

2-32

Figure 3

Figure 4

The condenser has a total of 71,328, 7/8 inch, 22 BWG titanium tubes. With a project of this
size, it is imperative that the turnkey package, supplied by the contractor, is complete and
includes:
Equipment:
oil free compressor(s) with aftercooler(s)
desiccant wheel dehumidification units with in-line booster heaters
dust collection units

2-33
Equipment: (cont.)
ultra high-pressure water jetting unit to remove existing coating and decontaminate the
tubesheet and tube-ends
abrasive blast equipment
three sizes of protective abrasive blasting plugs
three sizes of form plugs to keep coating from occluding tubes
storage/equipment trailer
small hand and specialty tools
lighting
ducting, air hose, etc.
QC equipment
Manpower:
full-time supervision
technicians trained to the contractors Application Qualification Program
application experience in nuclear stations
ability to gain unescorted personnel access to nuclear plants
thorough knowledge of the application procedure by all technicians to insure quality control
at each step in the application process
no sub-contractors used for surface preparation and application
Materials:
100% solid epoxy materials that are over-engineered for circulating water systems
suitable abrasive for cleaning and profiling
ASTM testing program to demonstrate properties of the coating materials
Application Procedures:
identifying temperature and humidity limitations
specifying required abrasive blast cleanliness and profile requirements
hold points for inspection
step-by-step procedure for application
mixing instructions for materials

2-34
specifying overcoat windows for multiple layers materials
Safety:
compliance with OSHA and plant requirements
confined space access
respiratory protection program
working on scaffolding
all required PPE
all required FME
The tubesheet cladding project was completed, on-schedule and at the firm price quoted. Several
unit-refueling outages provided the time for evaluation of the inlet tubesheet cladding. Its
durability was evident and reductions in sodium and chloride levels on the secondary steam side
of the condenser were achieved. It was decided that the four outlet tubesheets on Unit #1 would
be clad in May 2000. As a result of these efforts, and the effectiveness of the tubesheet cladding,
the main condenser leak rate dropped from approximately 20-30 gal/day and 2-3 ppb sodium, to
.2 gal/day with sodium values dropping to .05 ppb..
Based on the success of Unit #1, the same plan of action was detailed for Unit #2. In February
1995, four inlet tubesheets were clad. The four outlet tubesheets were clad in May 1999. During
the last Unit #2 refueling outage in March 2002, an extensive project to reduce main condenser
in-leakage to near zero gal/day was initiated. During this outage, a thorough steam side internal
inspection was conducted. The purpose of this inspection was to identify to identify steam
impingement areas, mechanical damage from miscellaneous debris, steam erosion of periphery
tubes, tubes at steam dumps and tube failures at support plates. An eddy current inspection
program was instituted for examination of 20 % of the 71,328 titanium tubes, with emphasis on
high-steam impingement areas. A condenser shell side hydrostatic test was conducted to
identify any remaining leakage by filling the condenser with water and dye solution. Tube leaks
were identified by black light and permanently plugged. Condensate sodium levels have
dropped dramatically to .098 ppb sodium.
Although the effects of FAC have been lessened as evidenced by lower iron values, additional
modifications are needed so that increased pH levels can be achieved. The station is presently
evaluating how to both bypass the Condensate Polishers to allow even further PH increases on
the steam side for piping protection, while maintaining the Condensate Polishers in a ready
standby state, to compensate for a condenser tube leak.
3-1
3
SESSION 3: MAINTENANCE TECHNOLOGY
Six Case Histories of High Reliability Coatings and Condenser Corrosion Problems
Michael J. Horn, James E. Mitchell
Plastocor, Inc.
The Use of 100% Solids Epoxy Coatings for Full Length Condenser Tube Linings
A. Wesley Langeland, Duromar, Inc.
Richard Kreiselmaier, Plastocor-international SA
Bruce Woodruff, Progress Energy
Heat Exchanger Tube Side Maintenance Repair vs. Replacement
Bruce W Schafer
Framatome ANP, Inc
Condenser Application and Maintenance Guide
Alan Grunsky, Project Manager
EPRI

Session 3: Maintenance Technology
3-3
SIX CASE HISTORIES OF HIGH RELIABILITY COATINGS
AND CONDENSER CORROSION PROBLEMS
Michael J. Horn, James E. Mitchell
Plastocor, Inc.
Hingham, Massachusetts 02043

Abstract
Corrosion, erosion, and mechanical forces have compromised the integrity of tubes, tubesheets,
and tube-to-tubesheet joints in steam surface condensers and other shell and tube heat
exchangers. Solutions include metallurgical re-engineering, cathodic protection, mechanical
rework, corrosion inhibitors, and coatings. While the dependability of coatings has been a
drawback, an extremely reliable coating approach has been developed. The integrity problems
of condensers at six power stations in the United States are outlined. In each case a high
reliability coatings system was successfully used. The elements of this approach are discussed.
Introduction
Corrosion and erosion of heat exchanger components in cooling water service is a major problem
in terms of: capital cost of replacement or maintenance, subsequent corrosion to primary plant
equipment due to in-leakage, and downtime costs for remediation.
Coatings and other plastic rebuilding compounds can both protect and restore heat exchanger
integrity. Long-term reliability of coatings has been a great concern. Engineered and applied
properly, coatings have proven to be a reliable long-term solution.
A steam surface condenser is a large, specialized, heat exchanger. While all references in this
paper are made to condensers, the same issues and solutions pertain to the balance-of-plant heat
exchange equipment. The case histories are drawn from condenser applications.
Causes for Loss of Tube, Tubesheet, and Tube-to-Tubesheet Joint Integrity
In conjunction with the tubes, the tubesheet acts both as a barrier between the cooling water
system and the condensate system, and as a structural member of the condenser. Tubesheets are
designed as heavy structures measuring an inch or more in thickness. Instances occur where
tubesheet corrosion has been so severe that its structural integrity has been compromised. Far
more common is that tubesheet corrosion first and foremost is a threat to tube-to-tubesheet joint
integrity, with the threat to condensate purity due to cooling water in-leakage. Generally, the
tube-to-tubesheet joint is a press fit where the tube is pressure rolled into the tubesheet.

3-4
In comparison, tube problems are the more common source of cooling water intrusion. However,
only one mode of tube failure is addressed here. And that is the selective degradation of tubes at
their inlet-end, usually termed: inlet-end erosion.
Compromise to tube, tubesheet, and tube-to-tubesheet joint integrity may occur via a variety of
mechanisms including:
General Erosion/Corrosion
While tube and tubesheet materials are designed to resist general corrosion, most do so by
forming a protective oxide layer. In service, cooling water turbulence in conjunction with
entrained solids can wipe this oxide layer away exposing the metal to subsequent cycles of
erosion/corrosion damage. General erosion/corrosion is usually an issue with tubes, particularly
at the inlet, and not the thicker tubesheets. This usually occurs with soft copper alloy tubing
where the cooling water has a high amount of suspended solids. Here the cooling water flow
selectively erode/corrodes the inlet tube ends where turbulence exists. Typically, the cooling
water flow smoothes out significantly after six to eight inches and the problem diminishes.
Galvanic Attack and Dealloying
A galvanic cell is created when dissimilar metals are placed in an electrolyte. The less noble
material becomes anodic and corrodes in the presence of the more noble (cathodic) material.
While the most active galvanic cells can be found with salt or brackish cooling water, significant
conductivity may occur in fresh water condensers especially those employing cooling towers in a
closed loop system. The trend towards upgrading to more corrosion resistant tubing (more
noble) such as stainless steel and titanium without changing or matching the tubesheets has led to
many instances of severe galvanic attack to the tubesheets. This attack may result in complete
metal loss or selective loss of a tubesheet alloy component. Selective loss is termed dealloying
and is further described by the constituent being attacked (i.e.. dezincification.
dealuminumification). Since condenser tubes are comparatively light structures, they are
designed to be as noble or more noble than the tubesheet. Hence, galvanic attack generally
affects the tubesheet and the tube-to-tubesheet joint.
General or Chemical Corrosion
The conductivity (ion content) or general oxidizing characteristics of cooling water is affected by
its chemical content due to pollution, mine runoff, concentration due to the use of closed loop
cooling systems and other factors.
Microbiologically Induced Corrosion (MIC)
Various microbiological organisms are corrosive. Generally these organisms must establish a
colony whereby corrosive metabolic by-products of that colony become concentrated and attack
the host metal. Generally a MIC habitat requires a foothold where water velocities are reduced.
Usually, the tube-to-tubesheet joint provides such a foothold. Pits, crevices, scaling, corrosion

3-5
deposits or other imperfections may also provide refuge. MIC generally occurs only with non-
copper bearing metals and generally appears like perforations or "worm holing" of the tubesheet
at the tube joint.
Crevice Corrosion
Depending on the metal, various additional corrosion mechanisms can occur in a crevice, or
other imperfection in the surface. These mechanisms (including MIC above) cause the crevice to
selectively get bigger and deeper. One common place for a crevice to occur is right at the tube
joint where galvanic attack initiated the crevice and is continuing to cause it. Such a growing
crevice may become a leak path as it undermines joint strength and integrity.
Mechanical Inadequacy
The mechanical strength of the tube-to-tubesheet joint may itself be substandard. Either under or
over rolling or other factors can leave the joint weak. When subjected to the thermal and
mechanical stresses of the condenser during operation, the tube can dislodge (even slightly),
breaking the joint strength and resulting in leakage.
Coatings as a Corrective Measure
A variety of corrective and remediation measures exist including tube/tubesheet replacement,
installation of cathodic protection, re-rolling of the tube ends, water treatments, and coatings.
Only coatings are addressed here.
Coatings have failed to perform because of: misapplication, physical damage in service,
workmen damage, cracking due to flexing or movement of tubesheets/waterboxes, or just the
long term inability of the coating to withstand the environment. In any event, it is commonly
feared that a flaw in a tubesheet coating may subject that area to selective galvanic attack (in the
presence of more noble tubing) with potentially catastrophic results.
A High Reliability Tubesheet Coating System
A coating system for tubesheets and tube-ends has been developed which has demonstrated the
high reliability required for long-term protection using high performance epoxy.
For tubesheets this process ( trademark: Plastocor) is essentially differentiated from conventional
coating approaches because of its thick-film nature. Precisely applied as a 200 mill molded
epoxy overlay to the tubesheet, the Plastocor process is more appropriately termed a "cladding"
(Figure 1). The first application was performed in 1963.

3-6

Figure 1
Cross-Section of Epoxy Tubesheet Cladding at 200 Mils
Such a thick application to the tubesheets provides long-term protection from corrosion, erosion,
and physical abuse. In addition, such a thick application confers the ability to not only
encapsulate and protect, but to rebuild/replace lost tubesheet metal with epoxy (Figure 2).

Figure 2
Schematic Cross-section of Rebuilt Tubesheet
While the primary purpose of the cladding is to protect the tubesheet and tube-to-tubesheet joint
seal, it also adds to joint strength. Tests using mock tubesheets have demonstrated up to 2400
pounds of additional push-out strength is added to the joint due to the adhesion and structural
strength of the epoxy materials. Additionally, the materials have demonstrated sufficient
flexibility to withstand service bending of the tubesheet.

3-7
The Tube-End ID Coating
For the tube-ends, the coating is necessarily applied thinly so as not to encroach on the tube ID.
The inlet coating is applied in addition to the tubesheet cladding so as to be anchored by the
cladding. Three coats are applied to the tube ID. Each coat is applied slightly deeper than the
first so as to achieve a feathered termination (Figure 3). Total thickness is six to nine mils.

Figure 3
Cross-section of Tube with Three Inlet Coats (exaggerated)
Elements of the High Reliability Approach
Epoxy Coating Materials
Over the past four decades, major advances in polymer chemistry have been made with respect
to coating technology. A variety of generic types of materials are readily available. Phenolics,
polyesters, vinylesters, urethanes, polyureas, fluoroelastomers, furans and epoxies have all had
major impacts on the coating industry. But the workhorse of them all continues to be the epoxy.
Epoxy technology has not stood still. Improvements in the resins, curing agents, modifiers,
fillers, chemically functional surface treatments, and other additives have resulted in dramatic
improvement in both the apply ability and serviceability of the materials.
In today's era of increased awareness of environmental issues, epoxies are finding new
applications in low or no VOC coatings, waterborne systems, and in materials that can meet the
new restrictions for hazardous waste. All of this while still maintaining its overall basic
strengths of rugged durability, outstanding adhesion, excellent chemical and temperature
resistance, and ease of application and repair which have made it such a useful polymer through
the years.

3-8
Methods of Application
Included in the development of the application specification is consideration of all aspects from
surface cleaning techniques, environmental control (humidity and temperature), material
qualification, mixing, application, intercoat windows, curing, and final testing. Hold points and
acceptance criteria are detailed. Central to the tubesheet coating system is the use of specialized
plugs which prevent the encroachment of the epoxies into the tube-ends, and which facilitates the
precise molding and buildup on the tubesheet surface. The inlet coating is applied by custom-
made high-speed rollers, which spin the materials onto the tube interior.
Application Labor
Experienced personnel generally apply the system on a turnkey basis. System procedures
include built-in quality control checks to insure the uniformity of the application. Turnkey
application not only serves to preserve consistent quality of materials, equipment, and
workmanship but also provides the benefits of single source responsibility. The training and
work history of the technicians is tracked.
Case Histories
The Plastocor System was introduced in the United States in 1984. One thousand (1000)
tubesheets, four hundred (400) waterboxes, seven (7) million tube-ends and thousands of feet of
piping have been successfully coated.
Case One: Severe Galvanic Attack of Carbon Steel Tubesheets
Materials of Construction
The condenser is constructed with 1" thick carbon steel tubesheets and 7/8" OD, 22 BWG, 304
stainless tubes. The waterboxes were uncoated carbon steel.
Cooling Water and Flow Configuration
The plant employs a closed loop cooling water system. Make-up water is drawn from a nearby
fresh water river.
Operating History
After ten years of operation, this 850 MW unit was experiencing severe loss of tubesheet
material due to general corrosion and galvanic attack. No cathodic protection had been in
operation. While the inlet tubesheets had areas where crevices up to 1/4" deep were discovered
in the carbon steel, the outlet tubesheets suffered more severe loss. On the outlet end the 304
stainless steel tubes protruded up to 1 1/2" from the sheet. Mud and other debris collected
between these protruding tubes creating a stagnant condition. The river mud was slightly acidic,

3-9
thereby enhancing the galvanic cell between the dissimilar metals. The resulting corrosion
caused wholesale loss of tubesheet ligament up to 1/2 " deep. The worst sections were over the
lower third of the sheet, which often remained submerged during unscheduled outages when the
condenser was not completely drained. In addition, crevices around the tubes were found to be
up to 1/4" deep. The combined result was that, in spots, up to 3/4" of the 1" tubesheet was lost.
The most severe metal loss was experienced on the final outlet tubesheet where the water
temperature is greatest in this triple pass design.
Solutions Considered
Given the severity of the metal loss, there was great concern over the integrity of the tubesheet
and the tube joints. It was decided that a coating must be applied that could rebuild the lost
metal and restore joint integrity as well as prevent any further corrosion. The Plastocor tubesheet
system was chosen. At the time of the coating installation, the waterboxes were also coated with
30 mils of epoxy. An impressed current cathodic protection system was installed as backup.
Service Experience
The Plastocor cladding was installed in November of 1984. Tube-to-tubesheet joint cooling
water in-leakage has not been observed. At last inspection the tubesheet cladding and waterbox
coating was found to be in excellent condition. It should be noted that the severity of tubesheet
corrosion was not completely realized until the cladding project was underway. Concern was
raised that the cladding could truly rebuild the tubesheets and represent a permanent fix. Capital
contingencies were made to replace the condenser if required. The service experience of the
cladding has been such that this contingency has been removed.
Maintenance Experience
At one point high pressure water was used to clean the tubes and in one area of one bundle the
lance operators allowed the water jets to dwell upon the tubesheet cladding and small divots
approximately 1/8" deep were created in the cladding. The damage did not extend to the
tubesheet metal and a repair has been scheduled as a precaution. As a result a maintenance
guideline has been issued so that operators of high-pressure water lances not allow the jets to
dwell on the cladding and that pressures above 5,000 not be employed.
Case Two: Tube-to-Tubesheet Joint Leakage
The condenser is constructed with 1" thick Muntz Metal tubesheets and 1-1/4" OD, 20 BWG,
90-10 CuNi tubes. The waterboxes are uncoated.

3-10
Fresh well water makeup on a closed loop system in the Southeast.
Operating History
After four years of operation, this 330 Mw unit experienced lost capacity during peak summer
load and increased demineralization costs due to poor condensate quality. An increase in low-
pressure turbine blade copper fouling was also noticed. There was no significant corrosion of the
tubesheets and the tubes were found not to have pitted so the cause of the contamination was
determined to be tube-to-tubesheet joint leakage.
Re-rolling the tubes was considered but determined to pose a greater risk than benefit as the
possibility of over-rolling existed. The Plastocor Tubesheet Cladding System was chosen as it
both sealed and fortified the tube-to-tubesheet joint.
Service Experience
The tubesheets were coated in November of 2000. At last inspection, the tubesheets were found
to be in excellent condition. Condensate quality was returned to normal and demineralization
costs were dramatically reduced. Copper deposits on the low-pressure turbine blades was
eliminated.
No maintenance of the coating has been required.
Case Three: Inlet-End Erosion and Tube-to-Tubesheet Joint Leakage
The condenser is constructed with 1" thick Muntz Metal tubesheets and 1" OD, 20 BWG, 90-10
CuNi tubes.
A once thru, two-pass unit using fresh river water located in the Southeast.

3-11
Operating History
After fourty-four years of operation this 100 MW unit began to experience tube failures due to
inlet-end erosion.
Re-tubing and a tubesheet cladding with an inlet coating were considered as alternatives. The
coating was chosen because the unit was frequently experiencing new tube-leaks (load
reduction), and the lead-time for retubing was unacceptable. In addition to the tubesheet
cladding, the first ten inches of the inlet tube-ends were lined by extending the Plastocor epoxy
down the tube at a thickness of 10 mils in order to prevent inlet end erosion.
Service Experience
The cladding and inlet coating were applied in November of 2001. No tube-to-tubesheet joint
leakage has been observed. Re-tubing of the unit has been postponed indefinitely.
No maintenance issues have been reported.
Case Four: Inlet-End Tube Erosion Resulting Tube Joint Inleakage
The condenser is constructed with 1" thick Muntz Metal tubesheets and 1" OD, 18 BWG,
admiralty brass tubes. The waterboxes are carbon steel and are epoxy lined.
The plant employs a once through cooling water system using brackish water in the Southeast
with high-entrained solids content.
Operating History
After ten years in service this 550 MW unit began to experience condensate contamination. The
inlet tube ends were severely eroded by the entrained solids in the cooling water.
Retubing was considered the primary option but financial and time constraints forced the
evaluation of alternative repairs. Both metal inserts and tube coatings were evaluated. Concern

3-12
with step erosion and tube ID restriction resulted in the selection of the coating option. At this
point both inserts and coatings were perceived as short-term solutions, which would be sufficient
to patch the unit until a retubing could be scheduled.
Service Experience
The cladding and inlet coating were applied in November of 1987. At last inspection the
tubesheet cladding and inlet coating was found to be in excellent condition. No tube-to-
tubesheet joint leakage has been observed. As a result the retubing has been indefinitely
postponed.
No maintenance of the coating has been required.
Case Five: Crevice Corrosion, Galvanic, and Microbiological Attack of Stainless Steel
Tubesheets
The condenser is constructed with 1" thick, 316L stainless tubesheets and 1 1/8" OD, 22 BWG,
294C tubes. The waterboxes are flake glass polyester lined carbon steel.
The plant employs a once through cooling water system using salt water in the Mid Atlantic
region.
Operating History
After eight years of commercial operation this 882 MW unit was experiencing failure of the
admiralty tubing. In 1982 the condenser was rebuilt using 316L stainless tubesheets and 294C
tubes. After two months of service the tubesheets were found to be suffering from crevice
corrosion and slight galvanic attack. Cathodic protection was installed to control the corrosion.
Subsequently it was determined that microbiological attack was also at play. By 1988 the
corrosion had thoroughly penetrated the tube joints so that over 1/2" of the joint had been
compromised. Furthermore, the corrosion, which had started with the tubesheet, was now
affecting the tubes such that they were being perforated from the outside in. However, little joint
in leakage was noticed. It was feared that further degradation of the joint would lead to failures.

3-13
Replacing the condenser was considered. The Plastocor cladding was chosen in conjunction
with the inlet coating with the expectation that it would seal the tubesheet and encapsulate the
perforated tube ends.
Operating Experience
The Cladding was installed in May of 1994. In 2001 the unit experienced a series of operational
anomalies including loss of cooling water (black-trip) and water-hammer. The tubesheet
cladding was damaged by these events and completely replaced in February of 2002.
Prior to re-coating replacement, no special maintenance was required.
Case Six: Galvanic Attack to Aluminum Bronze Tubesheets
The condenser is constructed with 1-1/4" aluminum bronze tubesheets and 7/8" OD, 22 BWG,
titanium tubes. The waterboxes are rubber-lined carbon steel.
The plant employs a once-through cooling water system using brackish river water.
Operating History
This 850 MW nuclear unit was first put in service in 1972 with 90-10 CuNi tubes. For a variety
of reasons including removal of copper carry over to the steam generator and other corrosion
issues the tubes were replaced with titanium in 1980. In concert with the retubing a protective
thin-film coating was applied to the tubesheets because of feared galvanic interaction. After
several years the tubesheet coating began to deteriorate. In 1991 it was decided to recoat the
unit.
An impressed current system and various coating options were evaluated. The Plastocor
cladding was selected because of its proven ultra-high reliability within the utility system and the
nuclear industry at large.

3-14
Operation Experience
The work was done in March of 1992. At last inspection the tubesheet cladding is in excellent
condition.
Workman damage associated with on-line maintenance has necessitated spot repairs of the
cladding.
Conclusions
Coatings and rebuilding compounds have been found to be useful in protecting cooling water
heat exchange systems from erosion and corrosion. Properly engineered and applied, coating
systems may serve to restore the integrity of heat exchange equipment as well as protect them
from further degradation.
Proper engineering should entail the evaluation and specification of four key elements:
1) Performance properties of the coatings
2) Application methodology and quality control
3) Qualifications and experience of the applicator
4) Thickness of the coating system
Often, the most under-emphasized of these elements is coating thickness. Properly applied,
thicker coatings will withstand service abuses longer than thinner coatings. Thick coating
systems in excess of 120 mils have been found to be extremely durable and are better classified
as "cladding" systems. Long-term performance in excess of twenty-five years has been
experienced for such cladding systems.
The six case histories presented here detail the usefulness of this approach in solving most
corrosion and erosion problems besetting steam surface condensers and other heat exchange
equipment.

3-15
THE USE OF 100% SOLIDS EPOXY COATINGS FOR FULL LENGTH
CONDENSER TUBE LININGS

A. Wesley Langeland
Duromar, Inc.
Richard Kreiselmaier
Plastocor-international SA
Bruce Woodruff
Progress Energy

Abstract
In an effort to control erosion, corrosion and fouling, the application of coating systems to the
inside of main steam condenser tubes has been the subject of some research over the past twenty
years. Recently, more interest has been generated as utilities try to extend the life or refurbish
older units where re-tubing may not be a cost effective option. Polymeric coatings, primarily
solvent based phenolics, epoxy phenolics, and polyesters, have been used to coat heat exchanger
tubes for years. Only recently, with the development of a new patented application system by
Plastocor-International, has the potential for the use of 100% solids epoxy systems become a
possibility.
Introduction
The interest in coating main steam condenser tubes dates back to work by the Japanese in the late
1970s, and possibly even before this. It was followed by research by the Italian utility ENEL,
who were looking for a solution to the fouling and corrosion problems on many of their coastal
and harbor units. Working with ENEL, an Italian researcher, Mario Pujsco, developed equipment
for applying a coating into the tubes by sending a spray lance down the tube and then extracting
it while spraying on a polymeric coating.
The early coatings used were epoxy phenolics and vinylesters, based on their application on heat
exchangers in the chemical industry. Because of equipment limitations, these systems were
heavily solvented in order to reduce the coating viscosity to a level which would enable the
material to be pumped down a long, thin hose and still be atomized at the nozzle. This lead to the
typical problems associated with solvent entrapment, poor film development, and non uniform
thickness. However, these coatings worked well enough to convince ENEL to coat several of
their seaside units.
Most of the early coating work centered on concerns over heat transfer. Coatings were evaluated
on the basis of effect on the overall heat transfer rate. Late in the 1980s, work done by Yusuf
Mussalli, and sponsored by the Empire State Electric Energy Research Corporation, indicated,

3-16
that tube coatings, if properly applied, do not reduce heat transfer appreciably in condenser
tubes.
1
This was followed up with additional work conducted at New England Powers
Manchester Street Station by Mussalli, Hager, & Tsou and sponsored by the Electric Power
Research Institute in 1989, which concluded essentially the same thing.
2
Later, work in 1993
conducted by Tsou and Woodruff at Florida Power and sponsored again by EPRI, centered on
the evaluation of generically different types of coating and there projected life from an economic
standpoint.
3
The results from this later work indicated the need for further work with respect to
coating development, equipment refinement and tube cleaning techniques. This paper presents
the progress which has been made in these three areas from a material standpoint.
Application Equipment
The newly developed and patented equipment for the application of coatings to the inside of
small diameter tubes and pipes overcomes several of the original application problems.
Most notable is the means of controlling the retraction speed of the spray lance in order to apply
a uniform thickness of the material throughout the entire length of the tube. Beyond this, spray
equipment, fittings and special hoses, which can operate at higher pressures, temperatures and
operating speeds, have been developed. This allows for the use of higher viscosity coatings
while maintaining optimal atomization levels. In order to reduce application times, a high level
of automatization has been added. This also adds to the reliability and reduces the chance for
operator error.
Surface Preparation
One of the most critical parameters to the long term performance of any coating is surface
preparation. This has been a particular burdensome problem when trying to clean old condenser
tubes. If too much pressure is used, a substantial risk of blowing holes in the tube inlet exists. If
the pressure is turned down, you run the risk of having no cleaning in the middle of the tube. A
variety of techniques have been employed to try and alleviate this concern, such as blasting from
both ends. This, however, greatly extends the time it takes to prepare the tubes, while still
overcoming problems with non-uniform cleaning.
The solution came from the reverse engineering of a technique used to eliminate cyclonic
spiraling in coal dust at burner injection points. This patented procedure effectively cleans a tube
uniformly over a distance of sixty feet, while removing no more material at the inlet than at the
outlet of the tube.
Design Criteria for a Main Steam Condenser Tube Coating
Because of the nature of utility power stations and the time which is available during outages for
maintenance work, working conditions as well as coating design parameters need to be
considered. For example, since most application work will require several weeks of time, in all
likelihood, other work will be taking place simultaneously. This means that solvent based or
flammable coating systems would be hazardous to work being carried out on the turbine deck or
in the condenser hotwell. In addition, coatings which lend themselves to easy application in a

3-17
shop environment or where elevated temperature curing is possible, would be impractical for use
on an existing condenser.
The major parameters used in the design of a coating for use on main steam condenser tubes are
listed below:
1. Resistance to the cooling medium, in most cases water and seawater
2. Resistance to condenser operating temperatures
3. Resistance to cathodic protection systems and galvanic corrosion
4. Resistance to erosion
5. Resistance to flexing and vibration, brittleness
6. Resistance to heat transfer
7. Resistance to fouling both biological and sedimentary
8. Resistance to film imperfections, thereby reducing the number of coats
9. Adhesion to the tube material and tubesheet
10. Cure times and temperatures.
11. Overcoat times
12. Disposal hazards

Beyond the operational concerns, these additional criteria are required for use with the patented
Plastocor tube lining equipment:
1. Long potlife
2. Low viscosity
3. Ease of atomization, surface tension
4. Low health risks
In order to meet as many of the above criteria as possible, a modified version of a 100% solids,
amido-amine cured epoxy was selected. The base material had been used for over ten years as a
manually applied coating for the inlet and outlet sections of condenser tubes to prevent erosion
from entrained solids in the cooling water.
Beyond the chemistry, a variety of conductive filler media was evaluated in an effort to improve
conductivity. The result was that in order to appreciably effect thermal conductivity in a thin
film coating, we needed to have almost 50% of the conductive media in the final film. A coating

3-18
having a consistency of a thin putty was required in order to achieve this level. For this reason,
we selected a silicon oxide filler, which gave us good overall erosion resistance in a thin film and
had a thermal conductivity of approximately 0.20 Btu-ft/ft2/Hr/F (ASTM E-228).
Application & Results
In April of 2001, approximately 7000, 90-10, 1 inch, Cu-Ni tubes were coated at a saltwater
cooled Florida utility at an average thickness of four (4) mils. Because of time and budget
constraints, only the upper sections of these condensers were coated as they were experiencing
the most problems. Operating results were not ideal as the first 3-6 feet of the inlet coating and 6-
9 feet of the out ends experienced blistering of the coating similar to what had been reported by
Nagata and Sato.
4
Although chlorides were checked for, it is believed that after testing,
additional chlorides entered the unit via the dehumidified air passing over seawater which
entered the outlet boxes at high tide. This blistering could also be aggravated by the cathodic
protection systems which could not be turned down because of the remaining 17,000 uncoated
tubes.
One month later in May of 2001, approximately 12,000, 1 inch, carbon steel tubes were coated at
a freshwater cooled European nuclear unit fitted with an online ball cleaning system. Operating
results after a year on this unit have thus far been good, with no reports of coating de-lamination
or inlet wear because of sponge balls.
Conclusion and Future Work
Although the present results from an operational standpoint, have been less than perfect, it
clearly points to the future potential of this alternative to re-tubing, especially when the long term
operation of the unit is in question. Ongoing development work is continuing, in an effort to
obtain higher temperature, more surface tolerant coatings with better thermal conductivity and
fouling resistance.

1
Yusuf G. Mussalli, C.R.B. Hoerger and F. Roma. Heat Transfer Resistances of Condenser Tube Coatings and
Liners.ASME-WAM, Anaheim, CA. December 1986.
2
Michael Hager, Yusuf Mussalli and John Tsou. Tube Coatings for Condensers and Heat Exchangers.
3
J.L. Tsou and B.N. Wooruff. Condenser Tube Protective Coating as an Alternative to Retubing.
4
Koji Nagata and Shiro Sato. Experiences of APF Condenser Tubes.Condenser Biofouling Control Symposium
The State-of-the-Art,Orlando, FL. June 1985.

3-19
HEAT EXCHANGER TUBE SIDE MAINTENANCE
REPAIR vs. REPLACEMENT
Bruce W Schafer
Framatome ANP, Inc.
155 Mill Ridge Road
Lynchburg, VA 24502
(434) 832-3360
bschafer@framatech.com
Abstract
The traditional method of repairing degraded tubes in shell-and-tube heat exchangers is to
remove the effected tubes from service by plugging. Since heat exchangers are designed with
excess heat transfer capability, approximately 10% of tubes can be plugged before performance
is affected. When the number of plugged tubes becomes excessive, heat exchanger efficiency is
lost, resulting in reduced power output, high system pressure drop, further heat exchanger
damage, or abnormal loads placed on other plant heat exchangers.
As an option to component retubing or replacement, repair methods, including tube sleeving and
tube expansion, have proven to be an effective method to repair defective tubes and keep the
existing heat exchanger in service. For the sleeving process, a new tube section is installed
inside the existing tube to bridge across the degraded area. Tube expansion is used to close off a
gap between the tube and the tubesheet or end plate (to eliminate a leak path) or between the tube
and tube support (to minimize vibration). While not all heat exchangers can be returned to their
original design condition by performing tube repairs, in some instances it may be possible to get
many more years of useful life out of a heat exchanger at a fraction the cost of replacement.
This paper presents options which the Plant Maintenance Engineer should consider in making
the repair versus replacement decision. This includes the repair options (sleeving and tube
expansion), other conditions within the heat exchanger, and the effect of tube repair on heat
exchanger performance.
Introduction
Traditionally, when maintenance is performed on shell-and-tube heat exchangers, the only
options considered when tube defects are found are to plug tubes and, when the number of plugs
became too great, replace the heat exchanger. The decision to replace the heat exchanger was
based on a number of factors. These included: the number of tubes plugged, the number of
forced outages due to tube damage (and the cost associated with replacing lost power and
repairing the damaged tubes), the impact that the plugged heat exchanger is having on the plant
(due to lost flow or heat transfer surface area), the rate at which tube plugging is occurring, the
availability of funds to replace the heat exchanger, and the expected life of the unit (how much
longer will the unit operate before retirement).

3-20
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). Since so many of
the failures were related to the tubing, the replacement of an entire heat exchanger due to damage
in one area is an expensive as well as a schedule and manpower intensive option.
The typical means for major heat exchanger repair included complete replacement, rebundling,
and retubing, as described below.
For the replacement option, the entire heat exchanger shell and tube bundle are replaced with
a new unit.
For rebundling, the shell is temporarily removed from the heat exchanger and the old tube
bundle, including, at a minimum, tubes, tube supports, and tubesheet, are removed. A new
tube bundle is inserted and the shell is welded back in place.
For retubing, either the shell (u-tube design) or tube side access cover (straight tubes) is
removed from the heat exchanger and the old tubes are removed from the bundle. New tubes
are then inserted and re-attached to the tubesheet (typically by either mechanical expansion,
welding, or both). In many instances, the existing shell side hardware is used as-is, although
some modifications may be made. Retubing is typically performed on straight tube heat
exchangers, such as condensers and coolers.
Since the 1970s, tube sleeving has been used to allow damaged tubes to remain in service. The
sleeves are installed by various means (roll, explosive, or hydraulic expansion, explosively
welded, or press-fit or epoxied in place) over the defective area of the tube. Through the use of
sleeving, which is a low-cost option to retubing, rebundling, or replacement, the useful life of a
heat exchanger can be economically extended. The decision to perform sleeving also can be
made with short notice as opposed to replacement (2-6 weeks compared with 18 months),
possibly allowing repairs to be performed the same outage that the damage is noted.
Tube expansion also can be performed to minimize or eliminate leakage within heat exchangers.
In the tubesheet, tubes can be re-expanded to strengthen the original tube-to-tubesheet joint,
reducing or eliminating leakage and prolonging the life of the heat exchanger. Expansions also
can be made deep within the tube to expand the tube into tube support plates and end plates.
These expansion can reduce tube-to-plate clearance for vibration control or, at end plates, to
minimize steam flow from the high to low pressure side of the plate.
Repair vs. Replace Factors To Consider
There are numerous factors to consider when deciding whether to repair the tubes in a heat
exchanger or to perform a larger repair scope and rebundle or replace the component. The
following factors should be considered when making the repair vs. replace decision.
The budget available for repair or replacement needs to be determined. Typically, the cost of
performing a substantial heat exchanger repair (consisting of plug removal, tube inspection,
tube expansion, 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.

If the heat exchanger is critical to plant operation (either from a safety, efficiency, or power

3-21
production standpoint) or is resulting in costly forced outages, it may be possible to justify a
repair to the unit in the near-term and a scheduled replacement when a longer outage can be
planned.

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 plug tubes, using 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.
The location and quantity of the tube defects need to be examined to decide if tube repair is
an option. Tube repair may be appropriate if the 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 are repaired. However, if the u-bend region of the tube is
damaged then tube repair is not possible. Also, it would not be possible to install a sleeve if
a large portion of the tube had damage but there was inadequate clearance for a long sleeve at
the tube end.
One of the more important items to consider when deciding whether a heat exchanger can be
repaired is the condition of the remainder of the 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.
The life expectancy of the power plant needs to be factored into the decision to repair or
replace a heat exchanger. 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.
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.

Along with the shorter outage duration, the site support required for repair is much less.
Usually, there are no shell or head modifications required since all work can usually be
performed through the manways and pass partition plates. Less repair equipment is required,
resulting in less space being needed in the area of the heat exchanger for setup and storage.

In addition, the time required to prepare for tube repair is much less than for replacement (2-
6 weeks compared with 18 months), allowing a decision on repair to be made just before, or
even during, an outage.

3-22
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.
If the heat exchanger is being replaced to eliminate detrimental materials in the cooling
system (i.e. copper in the condensate/feedwater system) then tube sleeving will not be
beneficial. The only practical solution would be to retube/rebundle/replace to change out the
tube material.
Heat Exchanger Repair Options
There have always been options available to either repair or replace heat exchanger tubes in the
event that tube leakage or degradation is present. The initial option, after the problem tubes have
been located (either through non-destructive examinations, such as eddy current testing, visual
inspections, or leak tests) is to plug the tube. Depending on the type of service and operating
pressures of the heat exchanger, various types of plugs are employed. These include tapered
fiber and metal pin plugs, rubber compression plugs, two piece ring and pin plugs, two piece
serrated ring and pin plugs (installed with a hydraulic cylinder), welded plugs, and explosively
welded plugs. In addition to the tube end plug, there also may be a stabilizer rod or cable that is
inserted into the tube to minimize future tube vibration damage.
At the beginning of the life of a heat exchanger, inserting a few plugs into damaged tubes has
little effect on the performance of the heat exchanger. However, if heat exchanger problems
continue, and the number of plugs increases significantly, it is possible that the heat exchanger
will eventually reach a point that it will not handle the full load that is placed on it. This is due
to a combination of loss of heat transfer area and the increased pressure drop. In addition, as the
number of plugged tubes increases, abnormal temperature conditions (either hot or cold spots)
may be set up in the heat exchanger. These conditions can result in an acceleration of tube
damage, creating a faster demise of the heat exchanger.
Once the number of plugs reaches a unacceptable level, the heat exchanger will need to be
repaired, replaced, or bypassed. However, bypassing the unit is usually not recommended, at
least for a long time period, since it will result in a loss of efficiency and heat transfer area.
Also, the heat load from the bypassed heat exchanger will be transferred to another heat
exchanger in the string, resulting in greater than normal operating flow rates and higher
degradation in that heater.
The following sections show the options that can be used to replace or repair the entire heat
exchanger or just the tubes.
Retubing
If the unit has straight tubes, good access, and the remaining components (shell, tube supports,
internal structural pieces) of the heat exchanger are in good shape, the tubes can be replaced.
The old tubes are removed from the unit and new ones, typically manufactured from an
improved material, are inserted, and then expanded, into place. Insertion of the new tubes is

3-23
shown in Figure 1. In addition to performing retubing to replace damaged tubes, retubing has
been performed to eliminate detrimental materials (such as copper from condenser tubes) to
minimize damage to other equipment within the plant (nuclear steam generators or fossil
boilers).

Figure 1
Condenser Retubing
Rebundling
Some heat exchangers are designed to be rebundled rather than replaced. For these units the
entire tube bundle, including tubes, tubesheet, and tube supports are replaced, as shown in
Figure 2. The original shell and any other internal structural pieces would be reused (although
any necessary internal repairs could be made when the shell was removed). The new tube
bundle can be manufactured to ensure that original design problems with the existing unit are
corrected. However, the same basic design must be maintained since the new bundle must fit
within the existing heat exchanger shell. Rebundling costs about 15-25% more than retubing (1).

Figure 2
Heat Exchanger Rebundling

3-24
Replacement
A third and typically widely used option is to replace the entire heat exchanger, as shown in
Figure 3. Full replacement allows alternate tube materials, changes in heat transfer area, and
structural changes to be employed, including added clearances in areas where erosion or other
problems may be occurring, to ensure that the current heat exchanger problems do not re-occur
in the future. However, the cost associated with a full replacement is the greatest of the three
options, about 5% more than for rebundling (1). In addition, there are no guarantees that the new
heat exchanger design will not have new, unanticipated problems.

Figure 3
Heat Exchanger Replacement
Sleeving
An alternate approach to retubing, rebundling, or replacement of a heat exchanger is to install
sleeves over the defective portions of the tubes The sleeve consists of a smaller diameter piece
of tubing that is inserted into the parent tube and positioned over the tube defects. After
insertion, each end of the sleeve is expanded into the parent tube material. These expansions
serve the dual function of structurally anchoring the sleeve into the tube and providing a leak
limiting path, allowing the sleeve to become the new pressure boundary for the tube. This means
that a sleeved tube can have a 100% through-wall indication and still remain in-service, since the
sleeve is now the new structural and pressure boundary. The installation of the sleeve into the
tube will allow the majority of the tubes heat transfer area and flow to be maintained.
If heat exchanger repair by sleeving is a possibility then a strategy needs to be used to prepare for
future repair. It may be cost effective to plug a quantity of tubes, per the non-destructive
examination results, each outage using a removable plug. When the quantity of plugged tubes
reaches a certain level the plugs can be removed and sleeves installed. Using this approach will

3-25
minimize the cost and time during each inspection outage while allowing the maximum tube
repair later in the heat exchangers life.
There are three types of sleeves that are installed into heat exchanger tubes. These are full
length, partial length structural, and partial length barrier sleeves. The three types are discussed
below. Figure 4 shows the sleeve layout.

Figure 4
Heat Exchanger Sleeve Designs
Full Length Sleeve
These sleeves are installed from one end of the tube to the other in 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%)
while minimizing flow pressure drop through the tube. After the full length expansion step,
shown in Figure 5, the sleeve ends are trimmed flush with the existing tube ends and the sleeve is
roll expanded into the tubesheet.
The full length sleeve is typically used in a condenser or cooling water heat exchanger when the
tubes have multiple defects along their length. Full length sleeving is an attractive option when a
relatively small percentage of the tubes require repair. Through sleeving, the majority of the
tube heat transfer area can be left in service, resulting in a heat exchanger that is close to its as-
designed condition.

3-26
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, making sleeving a cost-effective option
to return and keep tubes in service.

Figure 5
Full Length Sleeve Expansion
Partial Length Structural Sleeve
This type of sleeve is used to repair shorter defects in the tube. The sleeve can be installed
anywhere along the straight length of the tube. Various methods are used to expand the sleeve in
place. These include roll expansion (both in the tubesheet and in the freespan portion of the
tube), hydraulic expansion in the freespan portion of the tube, and full length expansion. These
expansion types are discussed below. The installation of a hydraulically expanded sleeve is
shown in Figure 6.
If one end of the sleeve is in the tubesheet, a torque-controlled roll expansion will be made.
This expansion is similar to the original tube-to-tubesheet roll. Freespan roll expansions are
made to either a torque controlled setting or to a diameter controlled hardstop setting.
Usually, freespan roll expansions are only used when the sleeve length is relatively short,
since it can be difficult to insert a roll expander deep into the tube. Both the tubesheet and
freespan roll expansion parameters are set so that they can provide both the structural and
leakage requirements for the sleeve.
For sleeves installed deep within the tube, a hydraulic expansion device is used to connect
the sleeve to the tube. The expander consists of multiple plastic bladders that are filled with
high pressure water. As the water pressure increases, the bladders expanded against the
inside of the sleeve, pushing the sleeve into the tube. The expansion process, which is
computer controlled, continues until either a preset volume of water or a preset pressure is
reached. At this point the sleeve is properly expanded and the bladders are depressurized.

3-27
Hydraulic expansions can be made anywhere along the tube length since the expander is
connected to flexible high pressure tubing and is not restricted by tube end access. The
expansion parameters are qualified to meet the proper structural and leakage requirements for
the sleeve.
Full length expansions are not usually used for structural or leak limiting purposes but
instead are used to improve heat transfer and flow through the sleeve and to close the annulus
between the sleeve and 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 to a preset pressure setting, expanding the sleeve into tight contact with the tube. After
the full length expansion is made, the ends of the sleeve are typically either roll or
hydraulically expanded to form the structural and leak limiting sleeve-to-tube joint.
Many times, the partial length structural sleeves are used to repair indications at one particular
area of the tube, such as wear damage at tube support locations, cracking in roll transitions, or
pitting indications at one discreet location along the tube length. Longer versions of these
sleeves also have been used to repair an entire damaged section of a heat exchanger, such as a
desuperheater or drain cooler section of a feedwater heater. Because of the wide variety of uses,
the sleeve length can range from as short as 1 foot to over 12 feet in length.
Qualification testing is performed on the structural sleeves to ensure that they can withstand the
design temperature and pressure conditions imposed on them. The test results must show that
the sleeve will be the new pressure boundary even with a 100% through-wall indication in the
parent tube. Sleeves of this type, using mechanical expansions (roll and hydraulic), have reliably
been in-service for more than 15 years.

Figure 6
Partial Length Hydraulically Expanded Structural Sleeve Installation

3-28
Partial Length Barrier Sleeve
These sleeves, also known as shields, are used at the ends of the tubes to act as a barrier to tube
end erosion. These sleeves are usually very thing, are not designed to act as a pressure boundary
or structural repair, and are installed in areas of high turbulence. The materials for these sleeves
are compatible with the existing tube material and may include plastic inserts. The sleeves are
either roll or hydraulic expanded or pressed or epoxied in place. If tube end erosion is occurring,
or is expected to occur, the use of these tube end sleeves will protect and prolong the life of the
parent tube, although over time tube erosion may begin to occur at the end of the sleeve. Many
heat exchanger tube ends have been protected with shields, significantly prolonging the life of
the tubes.
Items to Consider for Tube Repair
Prior to choosing to perform tube sleeving, the following factors should be considered.
The length, location, and quantity of tube defects that would require sleeving need to be
determined. If the defects are in one or a few short areas then either a single or a couple of
partial length sleeves could be used. However, if the defects are spaced throughout the
length of the tube, then the only option would be a full length sleeve.

The parent tube in the area where the sleeve will be expanded is to be defect free. This will
insure the highest sleeve-to-tube joint integrity. Also, the tube support designations must be
clearly identified to insure that the sleeve is installed at the correct location along the tube
length. This is especially true in areas where there may be skipped baffles and the tube only
touches every other support plate.
The condition of the remainder of the tube away from the sleevable defects needs to be
known. If there are u-bend defects that may require plugging then the tube should not be
sleeved. Sleeving is an option if the remainder of the tube is in good shape.
The space available at the tube end to insert a sleeve and its installation tooling needs to be
known, as shown in Figure 7. If a short, partial length sleeve is being used, the amount of
space required is not as critical, although there can still be access issues around the tubesheet
periphery for hemi-head channel covers and at pass partition plates. However, if a full length
sleeve is required, there will need to be a significant amount of clearance from the tubesheet
face.

3-29

Figure 7
Required Clearance for Sleeve Installation
Inspection records need to be reviewed to determine if there are any tube inside diameter
(ID) restrictions that would block the sleeve from being inserted to the target location. The
size of the eddy current probe used for the inspection, plus any other hardware that has been
inserted into the tube, can be used to help determine the tube ID access issues.
The post-sleeving tube inspection requirements need to be considered. Typically, the ability
to inspect the tube beyond a sleeve is not a significant issue. While the presence of the
sleeve reduces the inside diameter of the tube, which will result in the need for a smaller
inspection probe, the probe will remain large enough to detect pluggable tube indications
(usually greater than 40%), however small indications may go undetected.
As part of the post-sleeve inspection, the sleeve and its attachment to the tube should be
examined. There is no need to inspect the section of the parent tube between the sleeve
expansions since this is no longer part of the pressure boundary.
If tube cleaning is to be performed in the heat exchanger, then the type of sleeve to be
installed needs to be evaluated. If on-line cleaning is performed, the sleeve size cannot
restrict the passage of the balls or brushes. For off-line cleaning, the projectiles need to pass
through the sleeve without becoming stuck. Many sleeves that are installed in tubes that
require cleaning are full length expanded to ensure the best results for the cleaning
equipment.
If it appears that tube sleeving is possible, then information will be needed to ensure that the heat
exchanger is properly repaired. The following information is used when planning for sleeving.
Tube sleeving will need to be coordinated with eddy current inspection and plug removal.
If it is expected that sleeving may be performed, then it is important that the proper sleeve
material be purchased in advance of the job.
The sleeve material needs to be compatible with the heat exchanger parent tubing and with
the water chemistry within the heat exchanger. The galvanic corrosion potential between the
sleeve and tube needs to be determined. Also, effects of crevice corrosion between the

3-30
sleeve and tube, in the heat exchanger water chemistry, need to be considered to determine if
sleeving is a viable repair option.
The sleeve dimensions need to fit the heat exchanger operating and design conditions plus
any restrictions within the tube ID. The sleeve outside diameter (OD) is to be designed to fit
into the tube but must be long enough to limit the amount of sleeve expansion. The sleeve
wall thickness needs to be sized for the heat exchanger operating parameters, including any
ASME Code minimum wall thickness calculations, if needed. The sleeve length must be
long enough to span the expected tube defects but needs to be sized to fit any tube end
clearance restrictions.
Before installing sleeves into heat exchanger tubes, testing needs to be performed to set the
installation parameters. Depending on the type of sleeve being used, these tests may include
setting the rolling torque, hydraulic expansion constants, and full length expansion pressure.
In addition, depending on the application for the sleeve, there may be a need to do
qualification testing, which would consist of hydrostatic leak and pressure tests and
temperature and pressure cycling. These tests would verify that the expansion parameters
were set correctly for the sleeve application.
If a large quantity of sleeves are being installed, it may be necessary to calculate the heat
transfer and flow loss due to sleeving. These calculations will give a sleeve-to-plug ratio that
can be used to determine the expected improvement in heat exchanger performance after
sleeving is complete (and tubes have been returned to service, if applicable).
The sleeve may need to be full-length expanded based on the heat exchanger operating
environment. However, the production rates for sleeve installation are lower when full
length expansions are performed. While full length expansion is typically not needed in
many applications, such as most feedwater heaters, it should be considered for the following.
if tube ID cleaning needs to routinely be performed
if a long sleeve is being inserted that would severely restrict the tubes heat transfer or
flow
if the tube-to-sleeve crevice needs to be eliminated in a hostile water chemistry
environment
if there are large eddy current probe fill factor restrictions
Heat Exchanger Tube Expansion Repair
In addition to sleeving, it is possible to expand the tube to improve the heat exchanger
performance. These tube repairs can minimize further tube damage and maximize the useful life
of the heat exchanger. Two methods of tube expansion can be performed. One is to expand
deep within the tube to close off a leak path between the tube and the end plate. The other is to
re-expand the tube into the tubesheet to minimize tube-to-shell side leakage.
Tube-to-End Plate Expansion
In some heat exchangers, typically feedwater heaters, there are internal plates which separate one
zone of the heat exchanger from another (usually condensing [steam] from drain cooler [liquid]).

3-31
Due to the pressure differential across the plate, and the different temperatures and phases
between the two sections, it is important that leakage not occur through the plate. However, in
some feedwater heaters, the plate design is too thin, resulting in leakage of steam from the
condensing to the drain cooler zones, as shown in Figure 8. When this occurs there is erosion of
the end plate and tube vibration due to the high steam velocities and the steam condensing to
liquid in the drain cooler region. The vibration causes wear at the tube supports which can lead
to tube failure. The leakage of steam also increases the drain cooler temperature, resulting in a
less efficient heat exchanger.

Figure 8
End Plate Leakage in a Feedwater Heater
Expanding the tube can reduce the gap between the tube and the end plate. The expansion can
be performed using either a roll or hydraulic expander. Once the expander is in position the tube
is expanded until it contacts the end plate. An accurate expansion, which does not over-expand
the tube into the plate (the tube needs to be able to slide in the plate after expansion so that it
does not buckle during heatup/cooldown), needs to be performed. This can be achieved by using
a computer controlled hydraulic expansion that automatically shuts off the pressurization system
when it detects that the tube has contacted the plate.
After the tubes are expanded into the end plate, the steam flow is minimized or eliminated,
reducing the drain cooler temperatures and increases plant efficiency. Further tube damage, in
the form of tube wear and adjacent tubes impacting on one another, will be reduced to nearly
zero and the vibration operating stresses will be reduced significantly. The life of the heat
exchanger will be increased at a minimal cost as compared with replacement.

3-32
Tube-to-Tubesheet Expansion
In some heat exchanger designs, with a certain combination of materials, leaks develop between
the tube and tubesheet. In many low pressure units, the tube is only expanded into the tubesheet,
with no subsequent weld. Many of the leaks that occur in these units are the result of a
fabrication error and can be corrected by re-expanding the joint to the correct expansion size.
However, leakage occasionally occurs in high pressure heat exchangers, typically feedwater
heaters, even when the tubes have been welded to the tubesheet. The two prime causes of this
leakage are in areas where the original tube-to-tubesheet weld has either cracked or eroded due
to flow (in the case of soft materials, such as carbon steel) or where there is a crack in a tube-to-
tubesheet expansion transition.
For the first case it may be possible to re-expand the tube using a qualified roll expansion
process. The expansion would increase the contact pressure between the tube and tubesheet,
increasing the resistance to flow and decreasing or eliminating leakage. This process could
be performed on existing leaking tubes or preventatively on all tubes in the tubesheet.
If cracking is occurring at the original tube expansion transition it may be possible to
re-expand the tube deeper in the tubesheet (unless the cracking is occurring very close to the
shell side of the tubesheet). The tube would be expanded using a qualified roll expansion
process, to place the tube into tight contact with the tubesheet. This expansion would
increase the contact pressure between the tube and tubesheet, increasing the resistance to
flow and decreasing or eliminating leakage. This process could be performed either on
existing leaking tubes or preventatively on all tubes in the tubesheet.
Re-expanding tubes that either may be leaking or that could develop leaks in the future could
significantly extend the life of an otherwise good heat exchanger. By re-expanding the tubes,
forced outages can be avoided and damage from the high pressure water spraying on adjacent
tubes and on the shell will be eliminated. The cost to perform tube re-expansions will be
minimal when compared with the cost of replacement heat exchangers and the cost of forced
outages.
Items to Consider for Tube Expansion Repair
The following factors should be considered to determine if tube expansion is possible.
The portion of the tube to be expanded needs to be determined.
If leakage is occurring through the end plate, the expander will need to be long
enough to reach the end plate location. The tube should be expanded using a process,
such as hydraulic expansion, that will not lock the tube into the end plate. This
expansion will not only reduce leakage through the plate but also will minimize
future tube vibration due to the tight fit between the tube and plate.
If leakage is occurring within the tubesheet, due to either weld or tube cracking, a
re-expansion process may be used. This process, typically a roll expansion, will re-
expand the tube into the tubesheet to limit or eliminate leakage from the tube to the
shell side of the heat exchanger.

3-33
The condition of the remainder of the tube needs to be known. If there are cracks along the
entire tube length then re-expanding the tube alone will not result in an improvement in heat
exchanger performance.
The space available at the tube end to insert the expansion tooling needs to be known.
Usually either a roll or hydraulic expander will be used for this process. Unless a roll
expansion is being performed at the end plate, the usual repair tooling is relatively short,
although there can still be access issues around the tubesheet periphery for hemi-head
channel covers and at pass partition plates.
For tube end plate expansions, the eddy current inspection records need to be reviewed to
determine if there are any tube inside diameter restrictions that would block the expander
from being inserted to the end plate location. The size of the eddy current probe used for the
inspection, plus any other hardware that has been inserted into the tube, can be used to help
determine the tube ID access issues. The potential for any tube end restrictions, that might
limit tooling insertion into the tube, also need to be known so that tooling can be prepared to
eliminate the restriction.
If it appears that tube expansion is possible, then information will be needed to ensure that the
heat exchanger is properly repaired. The following information is used when planning for tube
expansion.
Tube expansion will need to be coordinated with eddy current inspection and plug removal.
The tube expander design (diameter and length) needs to be based on the requirements for
the expansion. Before performing tube expansions into heat exchanger tubes, testing needs
to be performed to set the tooling operating parameters. Depending on the type of expansion,
these tests may include setting the rolling torque for tubesheet re-expansions or setting the
hydraulic expansion constants for end plate expansions. In addition, for the tube-into-
tubesheet re-expansion process, qualification testing should be performed. This would
consist of hydrostatic leak and pressure tests and temperature and pressure cycling. These
tests would verify that the expansion parameters were set correctly for the tube re-
expansions.
Conclusions
The costs associated with heat exchanger replacement can be significant. These costs include the
new heat exchanger or tube bundle, the manpower required to remove the old and install the new
heat exchanger components, plant modifications to allow for the removal of the heat exchanger,
and the amount of outage time associated with replacement. In addition, the replacement of a
heat exchanger can adversely affect other work going on in the their vicinity. 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 either sleeving or tube expansion, 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 (2-6 weeks compared with 18 months), allowing a
decision on repair to be made just before, or even into, an outage.

3-34
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 also can 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. During this time, tube
repairs have economically extended the useful life of heat exchangers worldwide.
As the number of plugged tubes approaches the upper limits or if damage is consistently
occurring in one area of a heat exchanger, tube repair, through both sleeving and tube
expansions, should be considered to minimize future damage and extend the life of the heat
exchanger.
The following table shows the various heat exchanger repair options and the factors to be
considered when choosing each of the options. Note that the table contains selected criteria for
evaluating component repair versus replacement options. A final decision to implement a
particular option should be made on a case by case basis with proper weight given to all factors.
The information listed in this table is for relative comparison purposes only.
Table 1
Repair/Replacement Summary Table

Repair
Option

Application

On-Site Time to
Implement

Lead Time Required
to Implement

Longevity of
Selected Option
Component
Plus On-Site
Cost
Tube
Plugging
All tube defects, but limited
to ~10% of tubes before
affecting performance
Minimal time,
typically <1 week
Minimal time,
typically <1 week
Long term repair Minimal cost
Sleeving
Localized tube defects in
straight tube sections
Moderate time,
typically <3 weeks
Moderate time,
typically <1 month
Moderate to long
term repair
Less than 10%
of bundle or
heat
exchanger
replacement
Tube
Expansion
FWH end plate, tube wear
at tube supports, certain
cases of leakage at
tubesheet expansion joints
Minimal time,
typically <2 weeks
Minimal time,
typically <2 weeks
Moderate to long
term repair
Minimal cost
Tube Bundle
Replacement
All tube defects, limited to
those units designed for
replacement bundles
Maximum impact
time, typically
~3 weeks
Moderate to
extended time,
typically >4 months
Long term repair
Less than
component
replacement
Heat
Exchanger
Replacement
All tube and shell defects
Maximum impact
time, typically
~3 weeks
Maximum time
required to
implement, typically
>12 months
Long term repair
Highest cost to
implement

References
1. NMAC Feedwater Heater Maintenance Guide. Charlotte, NC: Electric Power Research
Institute, May 2002. EPRI 1003470.

3-35
EPRI

Filename.1
Photo
Condenser Application
and Maintenance Guide
1003088
agrunsky@epri.com
704-547-6056

3-36
Input on the 2000 EPRI NMAC survey and the 8/2000
NMAC Site Coordinators meeting requested a new
Condenser Guide. The intent of the guide was to:
Be a comprehensive guide for the Maintenance
Engineer at a Nuclear plant.
Replace an existing EPRI NMAC guide, ABCs of
Condenser Technology, published in 1994.
Emphasize the Reliability, Performance and
Maintenance Practices for the Condenser.
Be applicable for the Nuclear and Fossil units.
Overview

Overview - cont.
Technical Advisory Group members represented
CP&L, Dominion, Duke, Entergy, Exelon, PSE&G
and TVA utilities.
Vendors represented Conco, Expansion Seal
Technologies, Framatome, Heat Exchanger Institute
(HEI) and Plastocor.
A survey was sent to the Nuclear and Fossil plants for
Condenser Design, Materials, Cleaning Practices,
Water Treatment, Cathodic Protection, etc. The
results are tabulated in an appendix in the guide.

3-37
Contents
Introduction - Background (Why Guide Is Needed),
Approach, Guide Organization (Layout of Guide) and
the Use of Pop-outs (O&M Cost, Technical and
Human Performance).
Tutorial - Condenser Operation, Rankine Cycle,
Condenser Secondary Functions, Condenser Types (#
Compartments, # Passes, Transverse Versus Axial
Flow, Parallel or Series Design), Components (Shell,
Hotwell, Waterbox, Tubesheet, Tubes and Air
Removal Equipment).

Contents - cont.
Troubleshooting - Increased Condenser Pressure, Air
Binding Problems and Air Removal Equipment
Problems.
Performance - Heat Transfer, Condensing Duty, Heat
Transfer Coefficient, HEI Testing Method, ASME
Testing Method, Turbine Blade Effects, Performance
Monitoring, Software Tools (HEW-CA) and
Instrumentation.
Fouling - Macrofouling, Macrofouling Control
Technologies, Microfouling, Microfouling Chemical
Treatment, Fouling Monitor and Targeted
Chlorination with Fixed Nozzles.

3-38
Contents - cont.
Cleaning - Mechanical On-Line and Off-Line
Cleaning Systems and Chemical Cleaning.
Air/Water In-Leakage - Air In-Leakage Effects, Air
In-Leakage Detection Methods, Correcting Air In-
Leakage, Water In-Leakage Effects, Water In-Leakage
Detection Methods and Correcting Water In-Leakage.
Failure Modes - Industry Failure Data, Failure
Mechanisms and General Corrosion Prevention
Practices.

Contents - cont.
Condition Based Maintenance - Records, Periodic
Inspections, Preventive Maintenance and Non-
Destructive Examination.
Maintenance Repairs - Plugging Tubes, Tube Inserts,
Tube Sleeves, Tube End Coatings, Full Length Tube
Liners, Full Length Tube Coatings, Re-Expanding
Tubesheet Joints, Tubesheet Coating, Tube Staking for
Vibration, Waterbox Repairs, Tubesheet Repairs, Tube
Pulling and Miscellaneous Repairs.

3-39
Contents - cont.
Remaining Life, Materials and Constructability -
Remaining Life Assessment, Tube Material Selection,
Tubesheet Joints and Material Selection, Waterbox
and Shell Materials, Constructability Issues (Retubing)
and Rebundling.
References, Acronyms, Glossary.
Appendices - Survey Results, Mechanical Tube
Cleaning Procedure, Tube Plugging Procedures and
Pop-Out Summary.

Guide started in October 2000 with Final Report in
August 2001
EPRI Report Number 1003088
Available to NMAC and FMAC EPRI members
EPRI website - www.epri.comfor download
1-800-313-3774 to order hard copy
4-1
4
SESSION 4: PERFORMANCE IMPROVEMENT
A Condenser Success Story
Dave Leissner, Mirant
Richard Putman, Conco Consulting Corp.
On Understanding Condenser Pressure Saturation at Low Air Ingress
Joseph W. Harpster, Ph.D.
Intek, Inc.
Experiences With Steam Cycle Air Ingress at Alliant Energy Fossil Plants: Case Studies
Documenting the Effects of Air In-Leakage on Steam Plant Operations
Wesley A. Kaufman, P.E.
Alliant Energy IP&L
The Measurement of Condenser Losses Due to Fouling and Those Due to Air Ingress
Dr. Joseph W. Harpster, Intek Inc

Session 4: Performance Improvement
4-3
A CONDENSER SUCCESS STORY
Dave Leissner
Mirant
Chalk Point Generating Station
Aquasco, Md
david.leissner@mirant.com
Richard Putman
Conco Consulting Corp.
Verona, PA
reputman@concosystems.com

Abstract
The largest single factor that can affect turbine cycle efficiency within a generation station is the
heat transfer of the condenser. Depending on turbine design, poor condenser performance can
cost as much as 7% in extra fuel being consumed. Thus, maintaining condenser performance can
be of benefit not only economically to the plant but also to the environment. The factors affecting
condenser heat transfer are; tube-sheet macro fouling, tube deposits, air in-leakage and air
removal pump condition. This paper is a comprehensive look at Chalk Points Unit #2
condenser performance during the year 2000 and the corrective actions taken in 2001 to improve
unit backpressure by 2.5 in.Hg.
This condenser began the year 2000 with the actual backpressure at design conditions. This
paper details how, over the next twelve months, condenser performance deteriorated to the point
where the actual backpressure had risen to 2.5 in.Hg. above design. Because the unit was forced
to run on one circulating water pump for a period of eight weeks, factors such as tube-sheet
macro fouling, tube fouling deposits and silting was accelerated. The demand for MWs during
the summer, meant that the unit had to operate during the entire bio-fouling season without
cleaning the condenser water boxes. In addition to the fouling problems, air in-leakage began to
increase in September 2000 and its effect soon became as significant as the performance loss due
to fouling. The air in-leakage was caused by a failure of the cross over expansion joint in the
Low Pressure turbine combined with the rupture of the steam seal supply line to the gland seals.
These sources of air in-leakage were discovered only one month prior to the overhaul through
leak tests conducted with helium injection. Compounding the effect of the air in-leakage was the
fact that one air removal pump was due for a complete overhaul and the other pump was in need
of repair.

4-4
Introduction
The Chalk Point Generating Station is owned and operated by MIRANT. It is located on Eagle
Harbor Road in Aquasco, Maryland at the confluence of the Patuxent River and Swanson Creek
in the southeastern corner of Prince Georges County. The ambient air temperature varies from
15 F to 107 F. The site is approximately 15 ft above sea level.
Chalk Point Units 1 and 2 are identical B&W supercritical, once-through, double reheat,
regenerative cycle boilers and were converted to balanced draft design. Each unit is rated for
355 MWs. The furnace is of the open, water-cooled, dry bottom type. These units are designed
to burn pulverized coal through the front and rear wall burners. The boilers are designed for a
maximum rating of 2,500,000 LB/hr main steam flow at superheated outlet conditions of 1000 F
and 3,575 pig. At this rating, the reheat outlet conditions are 1050 F at 1020 pig and 1000 F at
315 pig for the first and second reheat stages, respectively. The condensers were built by
Worthington and are designed to condense 1,450,000 Lbs./Hr. of steam with a heat rejection of
1,435,000 BTU/Hr. At circulating water inlet temperatures of 59 degrees F the design Back-
pressure with a cleanliness factor of 85% is 1.11 Hg. The circulating water system consists of
two half capacity Worthington pumps rated at 125,000 GPM driven by GE motors rated at 600
HP. The circulating water system is counter-flow to the river with the discharge canal extending
over one mile upstream of the inlet structure to the condenser. The intake structure is
downstream of the discharge canal and brackish water from the Patuxent River is directed by the
inlet canal to the intake structure, which houses the pumps and inlet screens. The river water
temperatures vary from 32 degrees in the winter to 90 degrees F in the summer. The condenser
tubes were originally made of Aluminum Bronze material and were replaced with Titanium
tubes in 1987-88. There are four waterboxes to each condenser and each box is once through.
These units are also equipped with an Amertap online cleaning system.
Units 1 and 2 are each equipped with two Worthington reciprocating piston vacuum pumps
which are connected together by a common header leading to each of the air removal sections of
the condenser. The discharges from each of the air removal pumps are tied together and exit
through a common silencer before exiting the building into the atmosphere. The unit is designed
to run with only one vacuum pump in service. However, due to the air leakage caused by two
Low Pressure turbines and their gland seals, (even when the seals are in good condition), both
air removal pumps remain in service for efficient operation of the condenser.
Condenser performance monitoring on these units is accomplished using data obtained from the
Process Information system and additional testing conducted by vendors and plant personnel.
Initial Conditions June 2000
Unit #2 was on an extended outage during the spring of 2000. It was in March that all of the
condenser maintenance activities were performed such as bullet cleaning of the condenser tubes
and any air leakage repairs. This unit did not return to service until June and due to low
circulating water flow while the unit was off-line, much of the tube cleaning benefit was not
realized due to silt build up in the tubes. The importance of maintaining the proper circulating

4-5
water velocities through the condenser will be discussed later in this paper. The following were
the initial conditions of the condenser when the unit returned to service:
ACTUAL DESIGN
LOAD 355 MWS 355 MWS
CW INLET TEMP. 77.5 F 59 F
CW TEMP. RISE 14.9 F 14.0 F
BACKPRESSURE 2.6HG 2.2HG @ 77.5 F
HOTWELL O2 PPB * 2.8 PPB <5 PPB

*Notice that condenser air in-leakage was not a problem in June 2000.

Condenser Air Leakage
Hotwell Dissolved Oxygen
A very important part of condenser monitoring is the ability to determine the effectiveness of the
air removal capabilities of the condenser and the amount of air leakage that the condenser may
be experiencing. A necessary part of this monitoring is the dissolved oxygen analyzer that
determines the oxygen content of the condensate leaving the Hotwell. Chalk Point is fortunate
enough to have an on-line analyzer that is connected to the Plant Information system.
Below is a graph of the gradual increase in Hotwell dissolved oxygen beginning in Sept. 2000:

E2 HW O2
0
20
40
60
80
100
120
140
160
180
200
9
/
1
0
/
0
0
9
/
2
4
/
0
0
1
0
/
8
/
0
0
1
0
/
2
2
/
0
0
1
1
/
5
/
0
0
1
1
/
1
9
/
0
0
1
2
/
3
/
0
0
1
2
/
1
7
/
0
0
1
2
/
3
1
/
0
0
1
/
1
4
/
0
1
1
/
2
8
/
0
1
2
/
1
1
/
0
1
DATE
P
P
B

4-6
Air Removal Pump Testing
It was noticed that the level of Oxygen was increasing in the condensate from the hotwell. As a
result, tests were performed on the air removal pumps in order to determine their condition. The
test of the air removal pump involves shutting each pump off while the load remains constant
and recording the following:
Unit Backpressure
Circulating Water Inlet Temperature
Hotwell Temperature
Hotwell Oxygen Content
Air Removal Pump Motor Amps (Taken while both pumps are in service)
Fortunately, because of the many tests that have been conducted a baseline has been established
for this unit during these tests. With a tight condenser and when air removal pumps are in good
condition, when a pump is removed from service, the average increase in Hotwell O2 is 5 PPB
with a negligible increase in back-pressure and a one degree increase in hotwell temperature.
The following graphs indicate the vacuum response from a typical air removal pump test :

4-7

From the graphs above and the graphs on the following page, one would begin to make the
conclusion that the B pump is removing more non-condensable gas than the A pump. Another
tool that can be used to determine air removal pump condition is the trending of the motor amps
of the pumps. It has been our experience that as a one pump begins to deteriorate, the amps on
the motor decrease while the amps on the motor of the good pump increase. Trending of the air
removal pump motor amps, along with the air removal pump testing, has been an accurate
method of determining air removal pump condition.
The graphs below indicate typical hotwell temperature and dissolved oxygen response when air
removal pumps are taken off:
E2 PRIMARY VAC
a pump off b pump off
28.8
28.9
29
29.1
29.2
1
1
5
2
9
4
3
5
7
7
1
8
5
9
9
1
1
3
1
2
7
1
4
1
1
5
5
H
G

"
E2 SECONDARY VAC
a pump off b pump off
28.8
28.9
29
29.1
29.2
29.3
1
1
5
2
9
4
3
5
7
7
1
8
5
9
9
1
1
3
1
2
7
1
4
1
1
5
5
H
G

"

4-8

E 2 P R IM A R Y HW T E M P
a p u m p o f f b p u m p o f f
8 0
8 1
8 2
8 3
8 4
8 5
8 6
8 7
8 8
8 9
9 0
9 1
9 2
1
1
5
2
9
4
3
5
7
7
1
8
5
9
9
1
1
3
1
2
7
1
4
1
1
5
5
F
E1 SECONDARY HW T EMP
a p u m p o f f b p u m p o ff
8 0
8 1
8 2
8 3
8 4
8 5
8 6
8 7
8 8
8 9
9 0
9 1
9 2
1
1
5
2
9
4
3
5
7
7
1
8
5
9
9
1
1
3
1
2
7
1
4
1
1
5
5
F
E2 HW O2
a p u m p o ff b p u m p o ff
8
9
1 0
1 1
1 2
1 3
1 4
1 5
1 6
1 7
1 8
1 9
2 0
1
1
5
2
9
4
3
5
7
7
1
8
5
9
9
1
1
3
1
2
7
1
4
1
1
5
5
p
p
b

4-9
Helium Leak Testing
These units are periodically tested for air leakage using the conventional method of filling the
hotwell to the expansion joint with water. Filling the condenser with water to the expansion joint
is a very quick and easy method of finding air leakage below the expansion joint. During these
tests it is common to locate approximately a dozen small areas of air leakage. This supercritical
cycling unit runs approximately 100 days between boiler leaks, so that the opportunity to
perform this test occurs approximately four times a year when the unit is brought off-line.
Unfortunately however, this type of test does not identify any leaks that may exist on the turbine
deck. The best method of identifying leaks on the turbine deck is by use of helium injection. The
air leakage check during a short outage in the Fall failed to indicate any major leaks below the
expansion joints, therefore it was decided to contact Conco to perform a helium leak check to
help us locate and eliminate any air leakage on the turbine deck. Helium leak testing was
performed by Conco services in February of 2001. The following list indicates the location and
severity of leaks identified by the Helium leak test: (This test was conducted just prior to the
Spring overhaul.)
Turbine Deck
Secondary Low Pressure Cross-Over expansion Joint --- 60,000 divisions
Secondary Low Pressure West Gland Seal Packing --- 36,000 divisions
Secondary Low Pressure West Gland Seal Packing --- 36,000 divisions
Mezzanine Level
Secondary Condenser South Expansion Joint --- 18,000 divisions
(Helium may be carrying over to Cross-Over expansion joint)
Secondary Condenser North Expansion Joint --- 1,800 divisions
Basement
No leaks
Note: The term divisionsrefers to the level of response as indicated by the Helium leak
detector.
The helium testing confirmed the theory that the air leakage had to be coming from above the
expansion joints and located the leaks on the turbine deck. The major leaks of the cross-over
expansion joint and the steam seals on the low pressure turbine were definitely large enough to
cause problems with the heat transfer to the condenser tubes due to air leakage. The response
times of these leaks was almost instantaneous, it took less than 20 seconds for the helium to
arrive at the sample point just downstream of the silencer, which is located approximately 150
feet away from the air removal pumps.

4-10
Low Pressure Crossover Joints
Upon disassembly of the low pressure cross-over, significant damage to the bellows expansion
joint was found. The pictures below indicate the actual damage that produced an over-range
during the helium leak test of over 60,000 divisions.

4-11
Steam Seal Piping
As the dissolved oxygen began to increase in the hotwell, it was determined to be a serious
problem when the concentration exceeded 20 PPB, (parts per billion). The first assumption that
was made was that the steam seals on the low pressure turbine were beginning to deteriorate. As
a result of this assumption the pressure on the steam seal header was increased to compensate for
any additional clearance in the steam seals. However, when the pressure to the steam seals was
increased, there was no decrease in the dissolved oxygen in the hotwell. Since increasing the
steam seal pressure had no effect, it was theorized that the steam seal piping inside the LP
turbine may be leaking. This condition had occurred in cycling units #3 and #4, but had never
occurred in our base loaded units.
It was suggested at this time to increase the suction created by the steam packing exhauster in
order to determine any response in the dissolved oxygen. During this test as the steam packing
exhauster vacuum was being increased, the additional load on the exhauster motor caused it to
trip. Much to everyones surprise, the back-pressure on the condenser decreased along with the
dissolved oxygen content of the hotwell. This led us to believe that there was a serious problem
with the steam seal lines, causing us to place a high priority on the inspection of this piping
during the overhaul. The leakage rate of 16,000 divisions obtained during the helium testing
confirmed a significant problem in the steam seal area. Upon inspection, the steam seal piping
was found to be ruptured in several places; this not only allowed air to leak into the condenser,
but the steam leak also added to the heat load of the condenser. Unfortunately, no pictures taken
of the damaged area of the steam seal piping.
Biofouling
Low Circulating Water Flow Causes Silt Build-Up
Although the condenser was cleaned using metal cleaners, the unit was not immediately returned
to service due to the extended work performed during the overhaul. During this time, the
circulating water system needed to be placed into service to support the auxiliary equipment in
the plant. It was decided that in order to save on plant station service, that only one circulating
water pump would be run on the condenser being placed into service. As a consequence, the
tube water velocity fell to just 3.4 ft/sec, which is only half of its designed value. This condition
led to accelerated silt build-up within the tubes. The following graphs indicate the length of
time that each condenser was being operated with only one pump in service.

4-12

Continuous Summer Operation
After returning to service from the overhaul, the condenser remained in service until October
with no opportunity to clean the waterboxes of macro-fouling. The pictures below are an
indication of the amount of debris and accumulation of growth that units 1 and 2 see every
summer. Since being forced to reduce and replace the chlorine injection with non-chlorinated
circulating water treatment, the amount of macro-fouling due to garvia has increased
dramatically. The injection rates of the sodium hypochlorite are currently being increased and
under study in order to reduce the amount of garvia growth. As illustrated in the picture, when
we are not fighting garvia, leaves become a serious concern in the fall.
E2 CIRCULATING WATER
0
10
20
30
40
50
60
70
80
3
/
2
7
/
0
0
3
/
2
9
/
0
0
3
/
3
1
/
0
0
4
/
2
/
0
0
4
/
4
/
0
0
4
/
6
/
0
0
4
/
8
/
0
0
4
/
1
0
/
0
0
4
/
1
2
/
0
0
4
/
1
4
/
0
0
4
/
1
6
/
0
0
4
/
1
8
/
0
0
4
/
2
0
/
0
0
4
/
2
2
/
0
0
4
/
2
4
/
0
0
4
/
2
6
/
0
0
4
/
2
8
/
0
0
4
/
3
0
/
0
0
DAYS
P
U
M
P

A
M
P
S
A PMP
B PMP

4-13

Bullet Cleaning
Each year unit 1 and 2 condensers are cleaned using mechanical scrappers that are specifically
designed for the units tube ID and wall thickness. This activity is performed by Conco services
and is usually completed on a straight time basis in four to five days. There have been
occasions, however, when the cleanings needed to be completed in a single weekend and during

4-14
those times all 25,000 tubes were shot in less than 24 hours. Conco is our service provider of
choice due to their familiarity of our plant and their excellent track record.
The selection of tube cleaner design depends on the mineral analysis of the deposit. During the
cleaning process, tube deposit samples are taken and an ultimate analysis of the sample is
performed. Also, the total deposit weight is tracked and is used as a basis to determine if more
frequent cleanings of the condenser are required. When chlorine was used to control biological
growth, manganese was a large component of the tube deposit and due to the hardness of the
deposit, a stiff cleaner had to be used. However, a majority of our deposit is now silicon and iron
and as a result of this relatively soft deposit, we now use a cleaner with flexible blades.
During this last cleaning of unit #2 a video of the cleaning process was taken from inside the
tube with a specially designed camera. One of the surprising findings of the video was that the
garvia attached to the inside surface of the tube was not removed by a 300 psig water flush. (A
small portion of this video will be shown during the presentation of this paper)
Final Results
Condenser Air Leakage
The air leakage rate on Unit #2 continues to be under control and the air removal pumps appear
to be sufficient for removing non-condensable as evidenced by the following graph:

The chart below indicates the condenser performance before and after the overhaul. It should be
noted that in February prior to the overhaul that the performance delta from design was +2.7 Hg.
E 2 H W O 2
0
5
1 0
1 5
2 0
2 5
6
/
1
1
/
0
1
6
/
1
8
/
0
1
6
/
2
5
/
0
1
7
/
2
/
0
1
7
/
9
/
0
1
7
/
1
6
/
0
1
7
/
2
3
/
0
1
D A T E
P
P
B

4-15

The following table summarizes the results comparing June of 2000 to June of 2001:

DESIGN JUNE 2000
JUNE 2001 UNITS
LOAD 355 355 355 MWS
CW INLET TEMP 59 77.5 81.0 DEG F
CW OUTLET TEMP 14 14.9 14.0 DEG F
BACKPRESSURE 1.25 2.6 2.6 IN HG
BP DELTA FROM
DESIGN
.6 @ 77.5 F .2 @ 81.0 F * IN HG
HOTWELL O2 <5 2.8 3.4 Ppb

* Notice that the backpressure benefit from the cleaning is being maintained in 2001 because both circulating
water pumps remained on when the unit returned to service !

The result of this 2.7Hg back-pressure improvement, according to the back-pressure curves on
the turbine, is a reduction of unit heat rate by approximately 7%. This efficiency improvement
has already realized a fuel savings this year of $840,000 and a NOx emission savings of
$125,000. In fact, even though the circulating water temperatures are 30 degrees higher now in
July than what they were in February, the units heat rate remains 5% better.
Conclusions
The condition of the condenser is a major factor in determining unit heat rate and often
generation capacity. As a result, condenser performance has a tremendous impact on unit
operation economics. The performance monitoring of the condenser and routine removal of tube
fouling can significantly improve condenser heat transfer and greatly improve the efficiency of
the low pressure turbine. However, a decline in condenser performance may not be due to
fouling alone but also to either air ingress and/or the inability of the air removal system to
maintain the concentration of non-condensibles in the shell side of the condenser at an acceptable
level. Air ingress can occur in unexpected places while reduced performance of the vacuum
E 2 C O N D E N S E R
0 . 0 0
1 . 0 0
2 . 0 0
3 . 0 0
4 . 0 0
5 . 0 0
J
a
n
-
0
0
F
e
b
-
0
0
M
a
r
-
0
0
A
p
r
-
0
0
M
a
y
-
0
0
J
u
n
-
0
0
J
u
l
-
0
0
A
u
g
-
0
0
S
e
p
-
0
0
O
c
t
-
0
0
N
o
v
-
0
0
D
e
c
-
0
0
J
a
n
-
0
1
F
e
b
-
0
1
M
a
r
-
0
1
A
p
r
-
0
1
M
a
y
-
0
1
J
u
n
-
0
1
D A T E
I
N

H
G
0 . 0 0
0 . 5 0
1 . 0 0
1 . 5 0
2 . 0 0
2 . 5 0
3 . 0 0
B A C K P R E S S U R E
D E L T A F R O M D E S I G N

4-16
pumps can also be a factor. It has been shown that the condition of the latter can often be
determined from the results of quite simple tests.
Acknowledgements
We wish to thank the management of the Chalk Point Generating station for permission to
present this paper and would like to acknowledge the thoughtful suggestions contributed by
members of the staff of Conco Systems, Inc.

4-17
ON UNDERSTANDING CONDENSER PRESSURE SATURATION
AT LOW AIR INGRESS
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
When plotting condenser back pressure versus air in-leakage, it is often observed that there is a
low end region of pressure saturation. In describing this condenser behavior, argument is varied.
Common to most beliefs however, is that air ingress flow rate at the point where back pressure
begins to rise, is related to exhauster capacity.
Until recently there has not been a suitable condenser model to permit precise understanding of
the mechanisms responsible for this observation. This paper reviews this model and derives a
simple analytical explanation of this otherwise complex condenser phenomenon.

4-18
Introduction
A purpose of the condenser is to convert turbine exhaust steam to condensate at a sufficiently
low design pressure permitting an optimum amount of steam energy to be extracted by the
turbine. Higher back pressure at the turbine exhaust results in a load reduction, or an increase in
steam flow to maintain load; a higher heat rate is realized in either case. At lower than optimum
pressure, the exhaust steam can become choked in the turbine exhaust annulus with no net
increase in load, causing hotwell condensate to become subcooled, again increasing the heat rate.
It is reasonably observed that an increase in turbine back pressure of only 0.1 "HgA will cause a
change in plant heat rate of 0.2%.
One cause of increased back pressure on the turbine is Air In-Leakage (AIL). The amount of this
increase, referred to as excess back pressure, or correctable back pressure, over that of design, or
condenser pressure saturation value, is a function of the amount of air ingress. Further, there
exists a threshold value of AIL, A
o
, below which there is no excess back pressure and above
which condenser pressure rises as shown in Figure 1. It has been generally acknowledged that
this excess back pressure results from AIL being higher than exhauster capacity at a suction
pressure nearly equal to the condenser design pressure, P
D
. Lacking in explaining this noted
cause-and-effect relationship is a comprehensive understanding of the air/steam mixture
dynamics responsible for the observation.
A condenser model with analytical description was introduced by the author
(1,2)
in June and July
of 2001, presenting a new and unique understanding of condenser behavior. The model and
associated theory has been useful in understanding condenser performance parameters and
development of patented and patent pending condenser design features that eliminate air
binding
(3)
, reduce dissolved noncondensable gases in condensate
(4)
and improve condenser
performance, problem solving and operations.
Although the condenser excess pressure region was previously described
(2)
as a function of high
AIL, there was no analytical presentation of the dynamics of air/steam mixtures in the low AIL
pressure saturation region. This paper analyzes the latter region, extending it into the excess
back pressure region for completeness.

Model Basics
Air entering the condenser shell mixes with steam close to the leak source and is quickly
scavenged to the center of the tube bundle where it enters a small shrouded region of tubes in the
Air Removal Section (ARS). The air is then removed from the condenser through the vent line
by the exhauster as shown in Figure 2. The time taken for the air to reach the ARS, at low in-
leakage rates is generally less than one second and is dictated by the average tube bundle steam
velocity.
Air entering the bottom of the ARS is concentrated within the confines of the shroud which
lowers the heat transfer coefficient of ARS tubes in agreement with laboratory experiments
conducted by Henderson and Marchello
(5)
. These tubes represent about 1% of all tubes in the

4-19
bundle. The result of the lower heat transfer rate is that tube temperatures in the ARS are lower
than those outside the ARS. This lower temperature decreases the condensate surface
temperature located on these tubes which lowers the local water vapor pressure. The difference
between the higher steam saturation pressure outside the ARS and the lower water vapor
pressure in the ARS allows for a partial pressure of air to exist within the ARS space. The total
pressure in the ARS is about the same as the condenser pressure saturation value, particularly
near the inlet to the ARS. The temperature difference between ARS vapors, particularly near the
vent line entrance, and the condenser saturation temperature (steam temperature) is defined as
ARS subcooling. This subcooling is a feature of condenser design allowing the density of water
vapor leaving the ARS to be reduced, conserving condensate and enhancing the air removal
capacity of exhausters.
At higher AIL rates, subcooling in the ARS increases and reaches a limit when the ARS tubes
outlet circulating water temperature is reduced to the temperature of the inlet circulating water
temperature. In this case there is essentially no longer any condensation taking place on the ARS
tubes, particularly at the upper most row. Outside the ARS air has been scavenged into the
central core of the condenser expanding the subcooled region to tubes outside the ARS. Because
the region is subcooled, it no longer contains significant amounts of turbine exhaust steam but,
instead, lower temperature water vapor. This region, labeled with an Sin Figure 2, has been
designated a stagnantzone, devoid of rushing steam. All the tubes in this stagnant zone have
been essentially removed from the steam condensing process. The loss of these tubes requires an
increase in the logarithmic mean temperature difference to remove the heat of condensation from
the steam load over the remaining number of tubes and condensation surface area. As a result,
the saturation pressure of the condenser must rise. A 10 percent loss of tubes to a stagnant zone
is not much different from removing 10 percent of the tubes from the condenser, providing the
circulating water velocity remains the same. A factor , representing the fraction of remaining
active tubes in the condenser has been defined by Harpster.
(1, 6)
It is useful in distinguishing the
difference between heat transfer coefficient degradation caused by air and separately by tube
fouling.
An additional basis for analysis has been the consistent result from many plants that the water
vapor to air mass ratio, as measured with RheoVac
instruments
(7)
located in the condenser vent
line, is about 3 at the AIL threshold value noted for the onset of excess condenser back pressure.
The meaning here is that if AIL is increased above the threshold value, condenser tubes outside
but near the ARS entrance become increasingly surrounded by air, lowering the water vapor
temperature and forming a stagnant zone, which grows in size.
The resulting increased pressure within the condenser increases the density of air in the stagnant
zone and the ARS to the point that the exhauster, having a somewhat fixed volumetric flow rate,
can remove air in equilibrium with the AIL rate. When this occurs, there is air mass flow
balance and the stagnant zone size becomes stable.
Computations
Table 1 provides the results of theoretical computations on a condenser of the form depicted in
Figure 2. For simplicity, it has attached, an exhauster having a capacity, EC = 2000 ACFM. The

4-20
exhauster removes a mass flow rate of water vapor,
.
m
wv
, and air,
.
m
a
, in accordance with their
respective densities
wv
and
a
. The hotwell and steam temperature, T
HW
and T
s
respectively, are
equal (no hotwell subcooling) having a value of 108F under conditions of no, or low, AIL. The
low AIL saturation pressure, P
s,
is equal to the design pressure, P
D
, shown in Figure 1 having a
value of 2.45 "HgA as indicated in Table 1.

Table 1
Pressure Saturation and Excess Back Pressure Dependence on ARS and Stagnant Zone
Subcooling
%
Tubes
Lost
T
s

(F)
P
s

("HgA)
T
wv

(F)
p
wv

("HgA)
p
a

("HgA)
wv

(lb/ft
3
)
a

(lb/ft
3
)
.
m
wv

(lb/min)
.
m
a

(lb/min)
AIL
(SCFM)
.
m
wv
/
.
m
a
P
EX

("HgA)
0 108 2.45 108.0 2.450 0 .00357 0 7.14 0 0 0
0 108 2.45 107.5 2.415 .035 .00352 .00008 7.04 0.166 2.22 42.41 0
0 108 2.45 107.0 2.380 .070 .00347 .00016 6.94 0.328 4.38 21.16 0
0 108 2.45 106.0 2.310 .140 .00338 .00033 6.76 0.658 8.79 10.27 0
0 108 2.45 105.0 2.244 .206 .00329 .00049 6.58 0.972 12.98 6.400 0
0 108 2.45 104.0 2.180 .270 .00320 .00064 6.40 1.274 17.01 5.024 0
0 108 2.45 103.0 2.115 .335 .00311 .00079 6.22 1.584 21.15 3.927 0
0 108 2.45 102.0 2.053 .397 .00302 .00094 6.04 1.878 25.07 3.216 0
2 108.46 2.48 101.0 1.993 .491 .00294 .00116 5.87 2.324 31.03 2.527 0.03
4 108.96 2.52 100.0 1.933 .587 .00285 .00139 5.71 2.783 37.15 2.051 0.07
6 109.45 2.56 98.9 1.870 .690 .00277 .00164 5.54 3.280 43.79 1.690 0.11
11.1 110.83 2.66 96.3 1.728 .932 .00258 .00224 5.16 4.470 59.74 1.150 0.21
22.2 114.45 2.95 91.0 1.453 1.497 .00218 .00361 4.36 7.220 96.40 0.604 0.50
33.3 119.25 3.38 85.0 1.213 2.163 .00184 .00528 3.68 10.560 141 0.348 0.93
Conditions: Exhauster Capacity EC = 2000 ACFM, T
HW
= T
s
= 108F, P
sat
= 2.45 "HgA

The analysis can proceed, independent of steam load or number of tubes that make up the tube
bundle, until tubes start to become removed from the condensation process as a result of a
stagnant zone being initiated outside the ARS. At that point, T
s
must be determined from
individual tube steam load derived from the percent of tubes lost and a newly computed T
lm
and
TTD determined from circulating water temperature rise in tubes located in the active tube
region. The method is presented elsewhere.
(1)

The first column is an arbitrary selection of the number of tubes lost in the stagnant zone. The
next column is the selected or calculated steam temperature followed by the associated saturation
pressure or total condenser pressure. The fourth column is estimated vapor temperature values
caused by subcooling in the ARS or stagnant zone that will eventually result in a calculated value
of AIL, column 11.
With subcooling being T
s
- T
wv
, the lower temperature of T
wv
determines p
wv
from steam tables.
Subtracting p
wv
from P
s
provides p
a
, column 6. The density of water vapor,
wv
, also found from

4-21
steam tables, corresponds to T
wv
. The density of air is found from the ideal gas law for gas
mixtures given by:

a
= p
a

wv
/ 0.622p
wv
(1)
The mass flow rates of water vapor and air within the ARS or stagnant zone are determined from
the relationships:

.
m
wv
=
wv
EC

(2)
and

.
m
a
=
a
EC (3)

The total mass flow rate of gases being removed from the condenser is the sum of equations (2)
and (3).
The amount of AIL in SCFM units is determined by dividing
.
m
a
by the standard density of air
(0.0749 lb/ft) at atmospheric pressure and standard temperature. The mass ratio of water vapor
to air is determined either by the corresponding density ratio or the mass flow rate ratio.
The amount of subcooling is taken in one degree steps until the mass ratio of 3.2 is reached.
From that point on, the percent of tubes lost due to a stagnant zone was selected and the
necessary AIL rate to produce this loss was determined.
Results
Figure 3 presents the computational results of the theoretical model showing that an AIL
threshold value of 25 SCFM is expected for an exhauster having a total volumetric capacity of
2000 ACFM at the condenser design pressure P
D
= 2.45 "HgA. Above 25 SCFM, the condenser
will experience excess back pressure.
Figure 4 shows the total pressure and partial pressures of water vapor and air as a function of
AIL that can exist in the stagnant zone or the ARS as measured in the vent line. In the design
pressure saturation region, below 25 SCFM, it should be noted that p
wv
and p
a
exactly exchange,
with p
wv
having a negative slope and p
a
having a positive slope of equal value as AIL increases.
While p
a
continues to rise above the threshold value of AIL at the same slope, the water vapor
slope decreases, allowing condenser pressure to rise.
Also shown is the water vapor to air mass ratio versus AIL. These values are directly measured
by the RheoVac instrument representing the value obtained at the vent line entrance. The
theoretical values are shown by the solid line, representing a best fit to data of Table 1. The
large open circles represent actual data obtained in an operating condenser (JEA Northside Plant,
Unit #3). The model used for the analysis was a reconfiguration of the actual condenser,
arranged for analytical simplicity.

4-22
The total condenser pressure vs AIL theoretical curve also shows actual plant data for
comparison. The theoretical data points from Table 1 are represented by a best fit solid line and
again, the large open circles represent measured plant data.
Conclusions
It is believed that the condenser pressure saturation level for AIL values below a threshold value
has been described comprehensively for the first time. The method employed utilizes a new
condenser model first presented last year by Harpster
(1,2)
. The work reported here represents
another suitable result, supporting the validity of the model for its acceptance as being useful in
understanding condenser performance and to serve as a tool toward improving condenser design.
Many condenser design improvements have been identified
(3,4)
and are patented or patent
pending. It is now understood that subcooled regions of the condenser, one of which was
described herein, are responsible for all dissolved oxygen (DO) and other dissolved gases in
hotwell condensate, unless there exists a direct AIL into the condensate suction line or
contaminated condensate return draining directly into the hotwell. Removal of high DO
condensate at the source is part of these improvements.
Acknowledgments
The author wants to thank Mr. Fred Maner of JEA and the management of JEA for establishing a
cooperative research program whereby JEA benefits from utilization of an Intek supplied
RheoVac SENTRY condenser diagnostics system on Unit Number 3 and Intek is permitted to
perform tests in support of research goals. The work provided here and elsewhere would not
have been possible without this cooperative program.
Nomenclature

.
m
a
mass flow rate of air (lbs/min)

.
m
wv
mass flow rate of water vapor (lbs/min)
p
a
air partial pressure "HgA
p
wv
steam (wter vapor) partial pressure "HgA
P
s
condenser steam pressure "HgA
A
o
threshold air in-leak value
AIL air In-leakage, SCFM ft
3
/min |
STP

a
density of air lbs/ft
3

wv
density of water vapor lbs/ft
3

EC exhauster capacity, ACFM ft
3
/min |
operating

P
D
condenser design pressure or

4-23
pressure saturation value "HgA
T
s
steam temperature (F
T
lm
logarithmic mean temperature difference, (F
TTD terminal temperature difference (F
T
wv
water vapor temperature (F

References
1. Harpster, J.W. An Impact on Plant Performance from Advanced Instrumentationpresented
at the 44
th
Annual ISA POWID Industry Symposium; Orlando, FL (July 7-13, 2001).
[conference paper]
2. Harpster, J.W. On Understanding the Behavior of Non-condensables in the Shell Side of
Steam Surface Condenserspresented at the ASME International Joint Power Generation
Conference; New Orleans, LA (June 4-7, 2001). [conference paper]
3. Harpster, J.W. Reducing Dissolved Oxygen Under Conditions of High Air Ingress
presented at the ASME International Joint Power Generation Conference; Phoenix, AZ (June
24_26, 2002). [conference paper]
4. Harpster, J.W. Increased Availability from Improved Condenser Designpresented at the
Tenth International Conference on Nuclear Engineering; Arlington, VA (April 14-18, 2002).
[conference paper]
5. Henderson, C.L. and Marchello, J.M. "Film Condensation in the Presence of a Non-
Condensable Gas" Journal of Heat Transfer, Vol. 91, pp. 447-450, August 1969. [publication]
6. Harpster, J.W. Advancements on a Comprehensive Theory of Condenserspresented at the
45
th
Annual ISA POWID Industry Symposium; San Diego, CA (June 3-6, 2002). [conference
paper]
7. Harpster, J.W., Harpster, B.K. and Maner, F. Turbine Exhaust Excess Back Pressure
Reductionpresented at the FOMIS 38
th
Semiannual Conference - Optimizing Station
Performance; Clearwater Beach, FL (June 7-10, 1999). [conference paper]

4-24

Figure 1
Condenser Pressure vs. Air In-Leakage

Figure 2
A Typical Small, Simple Condenser
100 A
o
D
P
P
EX
2
4
0
P
r
e
s
s
u
r
e

(
"
H
g
A
)
Air In-Leakage (SCFM)
S
ARS
Feed
Water
Heater
AIL
To
Exhauster

4-25

Figure 3
Theoretical Condenser Back Pressure vs Air In-Leakage
2.44
2.46
2.48
2.50
2.52
2.54
2.56
2.58
2.60
2.62
2.64
2.66
2.68
0 10 20 30 40 50 60
C
o
n
d
e
n
s
e
r

P
r
e
s
s
u
r
e

(
"
H
g
A
)
Pex

4-26

Figure 4
Pressures and Mass Ratio vs AIL
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
4.0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
P
r
e
s
s
u
r
e

(
"
H
g
A
)
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
W
a
t
e
r

V
a
p
o
r

t
o

A
i
r

M
a
s
s

R
a
t
i
o
o
o
o
o
o = Condenser Measurement From
RheoVac and Plant Instruments
o
o
o
o
Condenser Pressure, P
S
Water Vapor Partial Pressure, p
wv
Air Partial Pressure, p
a

wv
/
a

4-27
EXPERIENCES WITH STEAM CYCLE AIR INGRESS AT
ALLIANT ENERGY FOSSIL PLANTS:
CASE STUDIES DOCUMENTING THE EFFECTS OF AIR IN-LEAKAGE ON
STEAM PLANT OPERATIONS
Wesley A. Kaufman, P.E.
Alliant Energy IP&L
P.O. Box 351
Cedar Rapids, IA 52406-0351
weskaufman@alliantenergy.com

Abstract
A variety of performance and operational problems result from air ingress to components
operating under a vacuum. This paper will describe experiences with identifying, locating and
correcting a variety of excessive air in-leakage incidents for condensers, turbines, and condensate
heaters.
The methods for discovering the source of in-leakage will be described along with their ability to
correctly identify the source of air in-leakage. Methods used at Alliant Energy include tracer gas
detection (helium and SF6), ultrasonic devices, infrared thermography, visual/streamers and
performance monitoring.
Several case studies have been accumulated over the past decade and will be presented to
document the effects, potential and actual costs associated with excessive air in-leakage. These
examples include:
Turbine/condenser expansion joint failure
Corroded thermowell on extraction steam line
Tube cleaning to confirm air in-leakage
Turbine inner shell casing leakage
Loss of seal water on air evactor
Vacuum relief valve/diaphragm leaking
Additionally, several temporary maintenance actions will be provided to minimize tramp air
online.

4-28
Effects of Air In-Leakage on Condenser Performance
Air is an effective insulator. The accumulation of air in the condenser decreases the effective
heat transfer surface by blanketing the condenser tubes and reducing the heat rejection capability
of the condensing steam. Air binding will also exist in areas of stagnant or reduced flow or if
there is insufficient air removal capacity. Air binding is generally not recognized by plant
operators, unless problems with the air removal system are known to exist.
The symptoms for excessive air leaks to the steam condenser are
increased exhaust steam temperature,
increased condenser terminal temperature difference,
increased condenser absolute pressure, and
high levels of dissolved oxygen in condensate.
For a properly operating condenser, some air in-leakage is normal through turbine shaft seals,
packing, and startup vents from feedwater heaters. When a significant leak appears, it is easier
to diagnose when the performance parameters of the condenser make an abrupt change. An
example may be after an outage, trip, or other transient operation.
The effects of air in-leakage are similar to other condenser related problems, such as obstructed
tube sheets or dirty condenser tubes. Because of the size of the vacuum boundary of the
condenser, which can include low pressure heaters, air ejection system, steam line drains, steam
traps, vents, extraction steam lines; it is usually easier to eliminate other variables as the source
of performance degradation.

Detecting Air In-Leakage Sources
Tracer Gases
Once air in-leakage is deemed to be higher than normal or excessive, cycle isolation becomes the
primary goal. If large surface area components or systems can be eliminated from the search, it
becomes easier to narrow the list of suspect areas for testing. Tracer gas techniques using sulfur
hexafluoride (SF6) or helium are best used to check large areas or plant sections. These systems
have evolved from cart-mounted units to smaller, more manageable and reliable units. Alliant
Energy has achieved positive results with these two types of tracer gas systems. We have used
an SAIC (now part of Conco) SF6 system and a Varian helium detection system. Both systems
have their merits, but the major criteria for successful tracer gas testing is ease of use of the test
system.
The need for pure hydrogen for the SF6 system was offset by that systems capability to provide
time stamped, hard copy output using a strip chart recorder. The SF6 gas dispensing system we
used was also easier to transport around plant equipment. The helium test kit is easier to setup

4-29
and calibrate, more compact and reliable, but the helium dispensing unit is heavier and bulky.
The output display for the helium monitoring system requires test personnel to constantly
monitor the LCD display and manually log responses- time of day and results.
The best situation occurs when the source of leakage can be isolated to a single component or
area using a tracer gas. However, some areas contain multiple small air leaks, numerous
individual joints, and connections may not be accessible. Locating specific leaks in such
circumstances can be difficult. For instance, leaks that draw air through conduits created by
insulation and piping configurations become mysterious, non-repeatable and nebulous (see Case
Study IV). If insulation can be removed to expose physical joints, bolt holes, flanges, etc., the
chances for positively locating leaks dramatically increases.
For truly large leaks, the amount of air movement may also make detection by tracer gas
difficult. Excess tracer gas will migrate towards the leak, causing tracer gas to accumulate in the
condenser or at the point of measurement. The background levels of tracer gas increase over the
course of testing and reduce the effective sensitivity of the measuring instrument as the
environment becomes saturated.
Ultrasonic Listening Devices
Once an area of leakage is identified, ultrasonic detection is the second level of inspection.
Ultrasonic leak detection is based on amplification of high frequency noise generated by
turbulent flow. Most pressurized leaks can be easily located using an ultrasonic listening device.
Vacuum leaks tend to be harder to find using ultrasonic inspection techniques, since sound
waves travel line of sight. The sound level, measured in decibels (dB), will be higher on the
low pressure side of a leak, which is inside the condenser. On the high pressure, atmospheric
side of the leak, air is moving at a slower bulk velocity as it converges at the suction point of the
leak.
Blocking techniques are usually required. Blocking is a term that describes shielding the suspect
area under investigation as much as possible from background noise. Nearly any cloth can be
used to insulate areas from neighboring steam leaks, bearings, motor noise, etc. Once a leak is
identified, ultrasonic devices are the easiest tool for verifying proper corrective maintenance has
been performed.
A drawback for ultrasonic detection is sound saturation over extended distances due to
reflections. In such cases, a telescoping extension can be made for the ultrasonic device to
extend the reach and capability of the machine and focus the sound sensitive microphone. At
Alliant Energy, we built a simple probe using 1and 7/8PVC pipe with stiff Tygon tubing
taped at one end. At the other end of the pipe, a piece of tubing was taped to form a transition
piece to convey the measured sound to the acoustic amplifier of the ultrasonic probe. All parts
were acquired at a local hardware store for under $14.

4-30
Infrared Thermography
Infrared thermography has not been beneficial to this author in locating vacuum leaks.
Typically, the mass flow of ambient air moving toward a leak does not have sufficient mass to
cool the surface temperature of the steel pressure boundary. Also, the ambient temperatures in
turbine rooms connected openly to boiler areas can be higher than the saturation temperature
within the condenser, negating any possibility of cooling at the point of the leak.
Visual Indications
At some point, over time, excessive leakage will cause tracings of dust/dirt on equipment
surfaces which provides a visual indication of the leak location. Narrowing down the location
can be done by covering large areas with cloth, gasket material or paint until the measured
leakage drops significantly. The bolted waterbox flanges at one of our facilities has been coated
with an epoxy paint to create an additional seal.
For very large leaks, a very light, strong streamer of material can be tied to a pole and used in a
manner similar to that used by operators tracking down high pressure steam leaks. Silk has been
used since it is very strong, light and less susceptible to melting/burning than polyester blends.
Due to the tight quarters between the turbine-generator pedestal and condenser, the pole should
be light and stiff. Bamboo sticks are preferred, being lighter than metal stock and also non-
conductive.
Care should be taken to ensure that the streamer is tightly attached to the extension. If the
material is drawn in through a penetration, the net effect of its retrieval is to further open the
crevice by cleaning the accumulation of dirt/oil that may have formed. This will be seen as a
further increase in DO levels or a decrease in condenser vacuum. The additional effect of
contaminated bulk material on the condensate pumps and piping cannot be assumed to be
beneficial.
Temporary Modification/Correction
For small diameter leaks at low temperature, almost any pliable material will have a positive
effect on reducing air in-leakage. To verify leaks, duct tape is suitable as a temporary packing
unless the size of the hole exceeds the open area between the fiber reinforcement of the tape. For
larger leaks, the use of rags, leather or gasket material can be taped in place or condenser
vacuum may be sufficient to hold them in place. If the unit cannot be shutdown for repair, there
are numerous room temperature vulcanized (RTV) sealants that will adhere to surfaces and
maintain an improved seal for days to months. Additional torquing of bolts has generally been
frowned upon due to the risk of bolt, flange or gasket failure.
Case Studies: Effects of Condenser Air In-Leakage
The following case studies show how different technologies were applied to locate air in-leakage
under several different plant configurations.

4-31
Case Study I: Turbine/Condenser Expansion Joint Failure
Location/Unit: Kapp Unit 2 200MW Westinghouse, Single Reheat
Background: Mississippi River is ultimate heat sink, once through cooling, admiralty condenser
tubes. Unit was experiencing secondary symptoms of high air in-leakage, running with high
dissolved oxygen over an extended period of time. This was increasing their boiler tube failure
rate, requiring high chemical usage, and resulting in condenser operation above 2.4 in.HgA after
condenser tube cleaning.
Testing: Air in-leakage tests using SF6 and Helium showed very little indications around
valves, fittings, pump shafts. Low pressure turbine shaft areas showed tracer gas response at the
high pressure turbine end and higher, but inconsistent responses at generator end of the low
pressure rotor- swinging two orders of magnitude. Visual inspection showed an accumulation of
dirt on the vertical wall of the condenser neck and air movement. Tracer gas technique was not
able to determine if shaft seal was the source. During a maintenance outage, a visual inspection
of the turbine expansion joint was performed inside the condenser and no cracks or breakage was
reported. An ultrasonic test was done to assist in locating the gas response. Due to the tight
location between the concrete turbine pedestal and condenser- 8inches- a telescoping stethoscope
was constructed from plastic tubing and duct tape. Ultrasonic readings were recorded along the
dogbone at 1-foot intervals, detecting a noticeable sound level increase 8ft from the edge. A
bamboo stick was used with a strip of silk to determine if the leak was significant. The cloth was
drawn in to the leak/condenser and difficult to remove. A larger cloth was used and drawn
completely off the bamboo rod.
Results: Backpressure immediately dropped from 3.26 in.HgA to 2.48 in.HgA (see Figure 1).
Dissolved oxygen dropped from 40 to 29 ppm, and continued to decrease. Review of
construction prints revealed the presence of a steam shield running around the interior of the
expansion joint, which was mistaken for the actual joint (Figure 2). Removal of the steam shield
revealed 12-inch long crack in the expansion joint material. Reduction in backpressure resulted
in a 563 Btu/kwh reduction in station heat rate. Coal procurement was reduced by approximately
42,000 tons/year of coal, for a savings of $675,000 annually.
Case Study II: Corroded Thermowell on Extraction Steam Line
Location/Unit: Sixth Street Unit 7 20MW GE, Automatic Extraction
Background: At low loads, both circulating water pumps were required to hold condenser
vacuum below 4.0 in.HgA. Station has Fisher ProVOX DCS, but limited data historian
capabilities. Operator experience was that backpressure had increased over a short period of
time, and vacuum was difficult to achieve on startup. Normal operation was requiring more and
more circulating water to maintain vacuum.
Testing: First thing was to review available hourly log data to determine how long problem had
been present. Unit responds to export steam demands and dispatch, so steady state data was not
available for all operating loads. To verify pump capability, the pumps were switched one at a

4-32
time with no effect on backpressure. Conclusion was that both pumps were supplying similar
flows, could not determine whether performance was degraded, but pumps did not appear to be
the problem. Tracer gas measuring system was not available, so ultrasonic gun was used to
listen for leaks. Prior to leak-checking, one of the operators remarked that the TTD on 11
th
stage
heater was high, over 10F, when it was usually less than 3F. Listening ultrasonically around
the 11
th
stage heater, the extraction line at the last elbow before heater shell flange was very
noisy. Closer inspection revealed the noise was coming from an empty thermowell on the
extraction line. To check whether the thermowell was the source of the disturbance, it was
covered with a glove and strong suction was present. Experimenting with the ultrasound device
and the thermowell showed that covering the thermowell greatly reduced the sound level. It also
caused the control room operators to experience temperature and pressure swings at the exhaust
hood.
Results: Although this turbine is small in capacity, it is dispatched as must run due to export
steam demands. Covering the thermowell reduced the condenser vacuum from 4.12 in.HgA to
0.97 in.HgA, and reduced the 11
th
stage heater TTD from 13F to 0F. Generator output
increased from 9.7 to 11 MW. Due to the warmer condensate temperature, heat rate dropped
1,558 Btu/kw-hr for a savings to the company of over $225,000 annually.

Main Steam Before
After
Units
Pressure 652 653 psig
Temperature 758 759 degF
Enthalpy 1380.9 1381.4btu/lb
Flow 260 259 kpph
Extraction
Enthalpy 1318.0 1315.2btu/lb
Flow 191 192 kpph
Condensate
Enthalpy 112.0 145.0Btu/lb
Flow 69 67 kpph
Backpressure 4.12 0.97 in.HgA
Generator 9.72 11.00 MW
GTHR 10,244 8,686Btu/kw-hr
Case Study III: Tube Cleaning to Confirm Air In-Leakage
Location/Unit: Sixth Street Unit 4 20MW GE, Automatic Extraction
Background: Unit output has been decreasing over time, becoming hard to maintain LP exhaust
hood temperatures at low condensing section loads. Condenser tubes were mechanically cleaned
with no appreciable affect on condenser backpressure or operation. Scheduled overhauls
deferred by financial constraints and economy of size.

4-33
Testing: First thing was to review available log data to determine how long high backpressure
problem had been present. Unit responds to export steam and electric dispatch demands, so
steady state data was not available for all operating loads. Tracer gas testing was performed on
the unit in 1997 and showed three leaks, two of which were corrected. The third leak was the
low pressure end turbine shaft. Review of previous outage report (1994) indicated some
variances with water seal clearance, and excessive wear of water seal impellar. At the time, it
was deemed that the impellar was still suitable for operation. By 1996, a flow meter at the
discharge of the steam jet air ejector had been bypassed due to excessive flow. A twin unit was
scheduled for condenser tube cleaning, and this unit was also cleaned hoping to reduce
operational problems. Further air leakage tracer gas and ultrasonic inspections in 2000
uncovered a inch hole in a drain header to the condenser, similar in size to the leak
experienced on Unit 7. Sealing this hole had very little effect on condenser backpressure or
turbine cycle performance, although the effect was positive.
Results: Cleaning the condenser tubes removed an average 110 grams of deposit per tube,
consisting of calcium (85%) and silicon (15%). The effect on unit operation was mixed-
temperature rise of cooling water and condenser backpressure are reduced, but heat rate
increased. By process of elimination, air in-leakage was determined to be the problem. This is
supported by the small effect of sealing the drain header hole. Also, it is possible that leakage
exists in the low pressure extraction steam line, since the 11
th
stage line is under vacuum. At low
condensing section steam flows, the unit experiences high hood temperatures. It is possible that
windage from air in-leakage is increasing friction in the last stages of the turbine. A major
turbine overhaul has been economically deferred from 2002 budget.
11/13/1999 3/29/2000
Before After
Main Steam Cleaning Cleaning Units
Enthalpy 1375.8 1381.5 btu/lb
Flow 178 187 kpph
Extraction
Enthalpy 1295.4 1284.4 btu/lb
Flow 106 111 kpph
Condensate
Enthalpy 151.4 150.1 Btu/lb
Flow 72 76 kpph
CW Inlet 61 61 degF
CW Outlet 73 68
Vacuum 2.29 1.46 in. HgA
Generator 8.3 8.3 MW
GTHR 11,717 12,554 Btu/kw-hr

4-34
Case Study IV: Turbine Inner Shell Casing Leakage
Location/Unit: Sutherland Unit 3 80MW Allis Chalmers, Single Reheat
Background: Following startup after major turbine overhaul, operations noted high exhaust
flow from steam jet air ejectors, above 12 inwc and less than 3 inwc prior to outage.
Testing: Tracer gas inspection was performed on the unit with SF6 system. Some minor leaks
were detected, but none large enough to account for the high air flow rate at the discharge of the
steam jet air ejector. A large, non-repeatable response was recorded along the side of the double
flow low pressure turbine. The relief diaphragms and shaft seals were checked repeatedly. The
response on the strip chart was never the same. The response was never quick, but often of low
magnitude. The response was generally a gradual increase in recorded tracer gas concentration
with a slow decay rate, similar to detecting a saturated gas condition within the condenser or
around the measuring device.
It was finally noted that the highest response level occurred when the tracer gas was blown into
gaps in the insulation skirting around the low pressure turbine. While the unit was online, the
skirting was removed to expose the turbine pedestal and all instrument penetrations around the
lower turbine shell. The gradual response continued, and the tracer gas response was again most
pronounced when gas was blown under the center insulation skirt that band the center of the low
pressure turbine. When the insulation topping the LP shell was removed, two bolted manways
were exposed. At this point, the ultrasonic gun was used to listen around the hatches and definite
responses were present. To verify that leaks were present, duct tape was used to cover the bolt
holes and both noise levels and SJAE flow decreased. With tape on 1/3 of the bolts and around
the flange, SJAE flow dropped to 3.3 inwc. Due to the high temperature, the duct was removed
and high temperature RTV was used in conjunction with the ultrasonic device until no sound was
detected.
Results: Air inleakage in to the outer shell of the turbine entered the lowest pressure heater
extraction steam line, causing the heater to become air bound. Isolating the extraction steam line
improved operation of the heater, but did not effect the volume of air leakage. Sealing the air
leaks on the manway improved both heater operation and condenser efficiency.

Sutherland Unit 3 6-Aug-97
As-found As-left
Parameter 6:30 10:30 Units
SJAE Exhaust +12.7 0.9 in.H2O
Cond From 3E LPH 140 168
O
F
Condenser Pressure 1.98 2.09 in.HgA
M.S. Flow 607 598 kpph
Generator Output 78 79 MWgross
Heat Rate 9,086 8,844 Btu/kwhr
-242 Btu/kwhr
Annual Fuel Savings $120,395

4-35
Case Study V: Loss of Vacuum Pump Seal Water
Location/Unit: Burlington Unit 1 200MW GE, Single Reheat
Background: Operators have noted that the Nash vacuum pumps on the condenser did not
appear to be working. Unit has two vacuum pumps, one small and one large, that were both
required to maintain vacuum on condenser. Pumps scheduled for replacement, engineering
requested performance test to determine appropriate sizing.
Testing: To measure each pumps capability, they were run individually. Pump A was stopped
in order to test pump B. Prior to restarting pump A, pump was primed (refilled). After starting
pump A, condenser backpressure dropped significantly.
Results: Centrifugal vacuum pumps was void of working fluid. Properly filling pump resulted
in decrease of backpressure by 0.75 inHg (Figure 8).
Case Study VI: Leaking Rupture Disk on Low Pressure Heater
Location/Unit: Ottumwa Unit 1 675 MW GE, Single Reheat

Background: FWH 102 (2
nd
lowest pressure condensate heater) terminal temperature difference
was high, from 25 F to +40 F. Operators have noted the flash evaporator rupture disk has
leaked in the past. Online performance monitor showed this heaters Target TTD should be less
than 20 F.
Testing: Performance monitoring software was activated in July, 2001. Shift supervisors and
control room operators have been reviewing display screens and parameter values for point
verification. Shift supervisor noted deviation of feedwater heater 102 TTD, and checked
extraction line and vessel for integrity. Suspecting the rupture disk was leaking, covered the disk
with plastic sheeting.
Results: Within thirty minutes of placing the plastic sheeting, TTD dropped from 41F to 13F.
Condensate outlet temperature increased 25F and extraction flow to heater increased from 70
kpph to 130 kpph (see Figure 9). Dissolved oxygen in the condensate system also dropped
dramatically
Summary
The most effective tools for locating air inleakage are
tracer gases for large, broad areas, and
ultrasonic devices for localized areas of concern.
Identification of air in-leakage problems is difficult to monitor and quantify, but effects can
usually be reduced online due to the low temperature.

4-36
As cycle isolation improves, even small air leaks can have a major effect on unit
performance.
The key to a successful air in-leakage reduction effort is persistence.
References
Information on Varian HeliTest system available at www.varianinc.com
Information on FluoroTracer monitor available at www.concosystems.com
Information on UltraProbe ultrasonic test kits available at www.uesystems.com
Information on FLIR thermography equipment available at www.flir.com
Acknowledgements
The author would like to acknowledge the assistance of the following individuals:
Phong Nguyen, Thermal Performance Engineer, for assisting with condenser leak investigation
at the M.L.Kapp Generating Station and his air removal investigation at the Burlington
Generating Station; and
Kevin Brehm, Shift Supervisor, Ottumwa Generating Station, for investigation and
documentation of the flash evaporator rupture disk failure.

4-37

Figure 1
Kapp Station, Unit 2. Effect of temporarily covering expansion joint tear with cloth.

Figure 2
Expansion joint side view showing steel liner on inside of condenser and mounting detail.

4-38

Figure 3
Sixth Street Unit 7. As Found conditions for broken thermowell on 11
th
stage extraction
line.

Figure 4
Sixth Street Unit 7. As-left conditions showing effect of plugging broken thermowell.

4-39

Figure 5
6
th
Street condenser 4, location of hole in drain header. Sealing this inch hole had
negligible effect on turbine/condenser operation.

Figure 6
Sutherland Unit 3 overview of LP skirting and center shell (casing) cover.

Generator
HP/IP

LP Casing
LP Insulation
Skirting
LP Shell
Cover

4-40

Figure 7
Sutherland Unit 3 with center shell cover removed and blanket insulation pulled away.
High temperature RTV gasket material (dark) contrasts with insulating blanket and casing.

Figure 8
Burlington Station Unit 1 Loss of priming seal water in centrifugal vacuum pump.

Burlington Generating Station
10,000
10,250
10,500
10,750
11,000
12:00 12:30 13:00 13:30 14:00 14:30
B
t
u
/
k
w
h
r
1.5
2.0
2.5
3.0
3.5
Net Unit Heat Rate Condenser Backpressure

4-41

Figure 9
Ottumwa Station, Unit 1. Temporary correction of bad flash evaporator rupture disk under
vacuum. Rupture disk covered with plastic sheet.

4-43
THE MEASUREMENT OF CONDENSER LOSSES DUE TO FOULING AND
THOSE DUE TO AIR INGRESS
Richard E. Putman Dr. Joseph W. Harpster
Technical Director President
Conco Consulting Corp. Intek Inc
Verona, PA, 15147 Westerville, OH, 43082

Abstract
The performance of a steam surface condenser is negatively affected not only by tube fouling but
also by the presence of excessive amounts of air within the shell side of the condenser. However,
because they have similar effects, it has been the custom to lump their costs together. Clearly, if
the contribution of each to operating cost could be estimated with some confidence, then
maintenance decisions would become focused on whether it is more important to clean the
condenser or to remove the source of an air inleakage.
There are two general methods for distinguishing between these two sources of condenser
performance degradation: (a) measure the resistance due to fouling and subtract this from the
total increase in tube thermal resistance to obtain the increase due to air ingress and (b) estimate
the change in tube thermal resistance due to air ingress and subtract this from the total increase in
thermal resistance to obtain the increase due to fouling.
For both methods, the frame of reference is an estimate of the total apparent increase in single-
tube heat transfer coefficient resulting from these two effects. This can be obtained by
calculating the present single-tube heat transfer coefficient using the standard Fourier equation,
and comparing it with the single-tube U-coefficient for a clean condenser operating under the
same load and cooling water conditions. In both cases, the condenser Performance Factor needs
to be applied
One approach to measurement method (a) is outlined in the new ASME Power Test Code
PTC.12.2-1998. A variation of this method is embodied in an EPRI/Bridger Scientific report in
which the flow through one of the tubes in each pair is also measured.
Measurement method (b) involves calibrating the degradation in performance due to air ingress
by injecting known quantities of air or nitrogen and also noting the reading of the flow meter
measuring the air removal rate. Subsequently the flow meter can be used to infer the degradation
due to air ingress based on the change in air removal rate. Interference with the precision of these
methods from air bindingand zones of stagnationare also discussed.

4-44
Introduction
Tube fouling as well as the presence of excessive amounts of air within the shell side of the
condenser both negatively affect the performance of a steam surface condenser. The thermal
resistance due to fouling reduces the overall tube heat transfer coefficient. Similarly, the
presence of air either reduces the shell side film heat transfer coefficient or, by completely
blanketing tubes, reduces the heat transfer area. Unfortunately, because they have similar effects,
their cost, measured as the increase in duty over that of a clean condenser, have had to be lumped
together. Clearly, if the contribution of each to operating cost could be estimated with some
confidence, maintenance decisions would become focused on whether it is more important to
clean the condenser or to remove the source of an air inleakage, or both.
To distinguish between the two sources of condenser performance degradation there are two
general methods: (a) measure the resistance due to fouling and subtract this from the total
increase in tube thermal resistance to obtain the increase due to air ingress and (b) estimate the
change in tube thermal resistance due to air ingress and subtract this from the total increase in
thermal resistance to obtain the increase due to fouling. Unfortunately, the task has been
hampered until recently by a lack of suitable instrumentation. The equipment associated with
these two methods, and now available, will be described; as well as the way the data is processed
in order to quantify separately the effect of each of these causes of condenser performance
degradation.
Condenser Performance Monitoring
Approaches (a) and (b) both require that the current increase in the single-tube thermal resistance
due to the combined effects of fouling and air ingress be known. This may be obtained by first
calculating the clean single-tube heat transfer coefficient U
ref
, which is a function of the sum of
the wall thermal resistance and the water and steam side film resistances, these being based on
the current operating conditions of cooling water inlet and outlet temperatures, tube water
velocity and back pressure. This becomes the reference. The principles involved were included
in both recent editions of the ASME Power Test Codes for Condensers
(1,2)
.
To obtain the equivalent single-tube value for the fouled conditions, the current value of the
effective heat transfer coefficient U
ref
is calculated using the well-known Fourier equation, as
outlined below, this then being modified by the value of the Performance Factor
(3)
that
corresponds to the present value of generated power. The increase in single-tube thermal
resistance may then be obtained by subtracting the fouled value from the clean value.
Clean Single-tube Heat Transfer Coefficient - U
ref
The thermal resistance to heat transfer of a clean tube consists of three major components:
i. Tube wall resistance
ii Water side film resistance
iii. Shell-side film resistance

4-45
Note that both (i) and (ii) are referred to the outer tube surface.
i. Tube Wall Resistance
The thermal resistance of the tube wall (R
w
) is calculated using the Kern
(4)
relationship:

(
=
i
o
m
o
w
d
d
k
d
R ln
24
(1)
ii. Water-side Film Resistance
The value of the water side film thermal resistance (R
t
) is calculated using the Rabas-Cane
correlation
(5)
:

(
=
i
o i
p
t
d
d
v
d
C k
. R
835 . 0
165 . 0
462 . 0 538 . 0 835 . 0
373 . 0
0450357 0
(2)
iii. Shell-side Film Resistance
The Nusselt factor (h
f
) is the condensate film heat transfer coefficient calculated from the
properties of water at the saturation temperature that corresponds to the compartmental
backpressure, the Nusselt equation being:

25 . 0
2 3
725 . 0
(
=
T) ( D
g k
h
o f
f
f

(3)

The shell side thermal resistance is1/h
f
.
Overall Heat Transfer Coefficient - U
ref
The overall heat transfer coefficient for a clean tube R
ref
can be calculated from the values of
resistances R
w
and R
t
and Nusselt factor h
f
as follows:

f t w
ref
h R R
U
/ 1
1
+ +
= (4)

Effective Condenser Heat Transfer Coefficient - U
eff
A steam surface condenser used in the Rankine Cycle is essentially a cross-flow heat exchanger.
The effective heat transfer coefficient (U
eff
) is computed from present steam and water
temperatures and cooling water flow rate and, by rearranging the well-known Fourier equation
for heat transfer, can be calculated from:

4-46

A*LMTD
Q
Ueff = (5)

in which
|
.
|
\
|
=
out v
in v
in out
T T
T T
T T
LMTD
ln

To calculate an accurate value of U
eff
requires knowledge of cooling water flow rate,
representative values of the inlet and outlet water temperatures, together with the compartmental
backpressure. For multi-compartment condensers, especially those operating at different back
pressures, this set of information is required for each compartment.
Performance Factor - PF
The design data sheet for a condenser designed in accordance with the Standards published by
the Heat Exchange Institute
(6)
contains information from which the HEI tube bundle heat transfer
coefficient U
HEI
may be calculated. It also contains information to calculate the value of the
effective heat transfer coefficient U
eff
as well as the design cleanliness factor CF
des
, this being:

HEI
eff
des
U
U
CF
* 100
= (6)

Thus the design cleanliness factor may be thought of as discounting the design HEI tube bundle
U-coefficient so as to reflect the operating conditions that will be experienced in practice.
When the thermal resistance method is used to calculate the single-tube heat transfer coefficient,
Tsou
(3)
recommends that the term performance factor be used in place of cleanliness factor. Thus,
Performance Factor may be calculated from:

ref
eff
U
U
PF 100 = (7)

A condenser designed with a cleanliness factor of 85% based on the Heat Exchange Institute
method
(6)
will have an equivalent performance factor of about 74% when the thermal resistance
method is used as the reference value. It should also be noted that the performance factor has
been found by Putman and Karg
(7)
to vary linearly with load, or:

PF = a
0
+ a
1
MW (8)

This must be taken into account when evaluating the condenser performance at partial load. A
typical plot of Performance Factor vs. load is shown in Figure 1.0 and it is necessary to establish
the relationship between Performance Factor (PF) and load when the condenser is clean. To
develop equation (8), the unit is run for an hour at steady state at various loads over the normal

4-47
operating load range and the Performance Factor calculated for each case. The essentially
straight-line relationship of equation (8) can then be obtained using regression analysis. Note
that, although displaced from one another, both cleanliness factor and performance factor follow
similar linear relationships with respect to generated power and have similar slopes.
Condenser Fouling Factor
Fouling factor R
fc
has been defined as the thermal resistance which can be attributed to fouling
and, when testing a single tube in a heat transfer rig, is calculated from the overall tube heat
transfer coefficient U
tot
from an expanded version of equation (4) as follows

fc f t w
tot
R h R R
U
+ + +
=
/ 1
1

To calculate condenser fouling resistance from the condenser effective heat transfer coefficient
U
eff
, it is necessary to perform the following transformation in order to convert the value of U
eff
to
the reference conditions:

fc
f
t w
eff
tot
R
h
R R
PF
U
U
+ + +
= =
1
1 * 100

from which
(
+ + =
f
t w
eff
fc
h
R R
U
PF
R
1
* 100
(11)

or
ref eff
fc
U U
PF
R
1
* 100
= (12)

where U
eff
is now a function of both fouling and excessive air ingress (if any) and may be
calculated using equation (5).
Thus the condenser fouling resistance R
fc
with reference to the clean single tube conditions can
be calculated from the condenser effective heat transfer coefficient modified by the performance
factor; minus the reciprocal of U
ref
.
If air is present having an air resistance R
a
and there is a tube fouling resistance R
f
, then the total
correctable resistance R
fc
may be written as:
R
fc
= R
a
+ R
f

Clearly, if R
fc
and R
f
are known, then: R
a
= R
fc
- R
f
(13)

Alternatively, if R
fc
and R
a
are known, then: R
f
= R
fc
- R
a
(14)

(10)
(9)

4-48
Methods of Measuring Fouling Resistance Alone (On a Tube Basis)
It is clear that the value of R
fc
includes the effect on heat transfer not only of fouling but also of
any air ingress. To distinguish between these two effects quantitatively, approach (a) above
would measure the fouling resistance directly in some manner and subtract this from R
fc
in order
to calculate the resistance that can be directly attributed to air ingress. Two methods for
estimating resistance due to fouling alone are available. The first of these methods was
developed as a part of the new ASME Performance Test Code for Steam Surface Condensers
(2)

and the other as an instrumentation system developed under the auspices of EPRI in conjunction
with Bridger Scientific
(8)
.
ASME Method for Estimating Tube Fouling Resistance
The Foreword to the new ASME Performance Test Code on Steam Surface Condensers
(2)
states
that to be certain that condenser performance results are not predestined, a mandatory
cleanliness test is now required by this Code.The method is illustrated in Figure 2.0, in which
the inlet and outlet temperatures are measured on sets of two adjacent tubes. One of the tubes in
each set remains in the as-found fouled condition while the neighboring tube has either been
cleaned or replaced with a new tube. It is claimed that both tubes in the pair experience identical
heat transfer conditions in terms of steam pressures, cooling water flow rates and velocities.
The Code suggests that the number of pairs of tubes selected for the fouling test be one per 2000
tubes per tube bundle: but not fewer than four pairs or more than 16 pairs per bundle. The pairs
are to be located at the approximate centroids of equal tube sectors within the tube bundle
pattern. However, pairs should not be placed within three tube rows of the bundle periphery.
To perform the fouling resistance test, not only are the outlet water temperatures to be measured
on each tube pair but also the common shell pressure, together with the cooling water inlet
temperature and flow. The fouling resistance of the pair in a single-compartment, which may
have one or two passes, is calculated as the difference between the heat transfer coefficients for
each tube in a pair, thus:
clean foul
f
U U
R
1 1
= (15)
Expanding and rearranging the Fourier equation (5) reduces to:

(
(
(
(
|
.
|
\
|
|
.
|
\
|
=
c out v
c in v
f out v
f in v p
f
T T
T T
T T
T T nwC
A
R
,
,
,
,
ln
1
ln
1
(16)

For `j pairs of tubes, the mean fouling resistance R
fmean
for the condenser tubes is given by:
=
=
=
j i
i
fi fmean R
j
R
1
1
(17)

4-49
As an on-line method for establishing fouling resistance for a short time after the one tube in
each pair has been cleaned, the method would seem to have possibilities. Further, it has been
assumed that the fouling resistance so calculated will not include the effects of any air ingress,
since the same effect will apply to both tubes in the pair. However, any fouling of the reference
(clean) tube will corrupt the values, as will errors in the estimation of cooling water mass flow
rate. Further, by measuring the inlet water temperature at only one point, any water stratification
will not be reflected in the results. However, inlet waterbox stratification is not anticipated to
have error within the same magnitude expected for outlet temperature measurements. This is
because temperature sensor measurements are strongly dependent on their location in the radial
thermal gradients of exiting water from individual tubes. Clearly, some may believe that a more
accurate result would also be obtained if the shell pressure adjacent to each pair were to be
measured rather than, again, measuring it at only a common point.
EPRI/Bridger Scientific Method for Estimating Tube Fouling Resistance
The principle of the method developed by Bridger Scientific under EPRI sponsorship for
estimating tube fouling resistance is illustrated in Figure 3.0. Tube pairs are again used but, in
this case, one of the pair is a tube with blanked off ends through which no water flows: while the
other, the fouled tube, not only has sensitive temperature measuring devices at both ends of the
tube but is also provided with a turbine type flow meter for accurate measurement of the water
flow rate through the tube. The blanked off tube is used to measure the mean shell temperature in
the vicinity of the fouled tube so that any vapor pressure loss through the tube bundles can be
allowed for. Several pairs of tubes are placed strategically throughout the tube bundle(s) so that a
mean fouling resistance can be estimated.
The interpretation of the data begins with substituting the water flow, water inlet and outlet
temperatures and steam temperature in the Fourier equation stated above in equation (5) and
calculating the effective heat transfer coefficient U
eff
. After having calculated the values of R
w
, R
t

and h
f
for the known tube operating conditions, the value of the fouling resistance R
f
can be
computed using equation (15). The mean fouling resistance R
fmean
can then be computed from
equation (17).
This apparatus avoids some of the criticisms that can be leveled at the method outlined in
PTC.12.2-1998
(2)
. The water flow and temperature rise is accurately measured: while the blank
tube allows the shell temperature in the locality of the fouled tube also to be measured with
precision. Unfortunately, the cost of the apparatus and its computer and instrumentation system
can be high but some economies might be possible if the calculations were executed within the
data acquisition system for the unit being monitored.
Calculating Fouling Resistance due to Air Ingress Using a Fouling Monitor
Once the values of R
fc
and R
fmean
are known, the apparent fouling resistance due to air ingress may
be calculated from:
R
a
= R
fc
- R
fmean
(18)

4-50
The Measurement of Thermal Resistance Due to Air Ingress
It has been shown, using a model and theory by Harpster,
(10, 11)
that air does not get trapped on
tubes throughout the condenser but forms a steam rich region and an air rich region within the
tube bundles. Referring to figure 4.0, the first of these regions is referred to as the Steam
Wind(SW) region and the second is referred to as the Stagnant(S) region
(10)
. It was also
shown, by Henderson, that if the mass ratio of steam vapor to air on a condensing tube is greater
than 300 the heat transfer coefficient is greater than 90% of its U
foul
condition. If this ratio falls
below 3, then the heat transfer coefficient will fall below 10% of its U
foul
value. In operating
condensers, the mass ratio can vary, practically, between 50,000 to 0.2 depending on location in
the tube bundle.
(11)

In a normal operating condenser having an exhauster removing air in equilibrium with the in-
leakage, sufficient to prevent condenser excess back pressure, the fraction of tubes essentially
unaffected by this air is greater than 96%. Further, air inleakage nearly 6 times the above pump
capacity value, sufficient to cause an additional 0.9"HgA in the measured condenser pressure,
will still have nearly 50% of its tubes, located in the outer regions of the tube bundle unaffected
in their measured heat transfer coefficient because the mass ratio in this region is greater than
1,000. These results are provided in the referenced literature.
(11)

Another perturbing feature about condensers explained by Harpster
(12, 13)
is air binding. This
phenomena results from condenser design that promotes steam flow completely surrounding a
tube bundle subsection and having no escape for scavenged air. Although these problems can be
overcome by design (patent pending) unawareness of their affects can give rise to a lack of
comfort regarding utilization of recorded data.
For these reasons, measurements using tube pairs should consider the above model result,
regardless of the array pattern recommended by the ASME test code. Figure 4 shows what
might be expected in a typical single shell, single pass condenser. The tube bundle consists of
three subsections separated by a vertical crevice between Subsection I and II where the air
removal section (ARS) vent line is placed. Another separation is caused by a horizontal
condensate tray above Subsection III. The condensate tray provides drainage of the condensate
from the above subsections to the sides of the tube bundle, from which condensate is allowed to
fall into the hotwell. These trays prevent inundation of tubes in the lower part of Subsection III.
The anticipated air bound (AB) regions, which, out of necessity grow in size and then collapse,
are shown in the tube bundle subsections at their most likely locations considering the bundle
configuration. Also shown is the "S" region having a high concentration of air. This "S" region
near the ARS is variable in size with air in-leakage and can be changed by admitting more or less
air (or an inert noncondensable gas like N
2
) at fixed flow rates into the condenser. Between each
adjustment, about 20 minutes is needed to establish equilibrium. A typical relationship between
the apparent increase in fouling resistance due to air ingress and the air removal rate is shown in
Figure 5.0.
Suggested tube pair areas to measure the effects of tube fouling without the impact of air are
shown as rectangular areas containing an "X" and labeled with "SW" indicating these areas are in
the "Steam Wind" region of the tube bundle.

4-51

Suggested tube pair areas for measuring the combined heat transfer coefficient of air and fouling
or just fouling are shown as square areas with an "X". These areas for a tight condenser i.e., for
air in-leakage below the excess back pressure threshold pairs, will measure the effect of tube
fouling only. With the addition of noncondensables, the stagnant zone will expand into these
areas and the pairs will record the effects of increased air concentration on the heat transfer
coefficient of these tubes.
Baselining and Controlling Air Inleakage
Following installation of the tube pairs the amount of background air in-leakage, water vapor to
air mass ratio and exhauster capacity for noncondensables must be measured. These are easily
determined using a Multi-Sensor Probe (MSP) measurement system, shown in Figure 6.0. This
system permits simultaneous measurement of air and water vapor flowing from the condenser in
each vent line penetrating the shell. Typically, there are more than one shell and these may have
the same or different pressures. Each vent line must be measured independently for air in-
leakage, since the amount of noncondensables are generally different in each line.
The total amount of background air should be well below the exhauster capacity and the
condenser pressure at the pressure saturation valve. If not, a leak search should be made and
leaks repaired before starting tests. Under this condition all tube pairs should be measured as a
baseline. It is expected that all pairs should have the same determined U
foul
values. If not these
values will serve as a bases for determining changes in measured heat transfer coefficients U
foul,air

upon introduction of air.
Air may be introduced into the condenser at any convenient location. It should be recognized
that this air will be scavenged by the steam to the closest ARS section. If uniform effects are
desired the air should be introduced on the turbine floor near the LP turbine exhaust annulus.
A convenient means to introduce air, or other gas, is to pass it through a rotameter adjusted for
atmospheric pressure at its inlet. If air is used, a control valve in the line between the top of the
rotameter and the shell is all that is needed. The plant air being drawn in through the bottom
opening of the rotameter should be free of steam and large amounts of dust or dirt to prevent
error in readings.
Estimating Tube Fouling Resistance from Air Removal Rate
Once the condenser has been calibrated to provide an estimate of the apparent increase in
combined fouling resistance due to air ingress R
a
, the current value of R
a
may be obtained from
the plot shown in Figure 5, the thermal resistance due to tube fouling then being estimated from:

R
f
= R
fc
- R
a
(19)

4-52
Distribution of Condenser Losses
Using a Newton-Raphson model of the condenser/turbogenerator subsystem, Putman and
Saxon
(9)
showed how total condenser losses in MBTU/h (Loss) can be calculated from the present
condenser duty minus the condenser duty calculated if the condenser were clean and operating
under the same cooling water inlet temperature and flow conditions and the same generated
power. If the fuel cost is $Cost ($/MBTU), then the distribution of these losses between fouling
and air ingress can be accomplished as follows:
$ Cost of Fouling:
fc
f
f
R
R
Loss Cost Loss * $ = (20)

$ Cost of Air Ingress:
fc
a
a
R
R
Loss Cost Loss * $ = (21)
Conclusions
Tube fouling and air ingress have a similar effect on condenser performance degradation.
However, methods for distinguishing between these two causes of performance degradation have
been restricted by the absence of suitable instrumentation. Three methods for quantifying the
contribution of these two sources of performance degradation are outlined as well as how they
can be converted to the equivalent economic loss.

4-53
Nomenclature

a
0
, a
1
= Constants in equation (8)
A = Total tube surface area for compartment ft
2

C
p
= Specific heat of water BTU/(lb.F)
CF = Cleanliness factor %
d
i
= Inside diameter of condenser tubes inches
d
o
= Outside diameter of condenser tubes inches
D
o
= Outside diameter of condenser tubes feet
g = Acceleration due to gravity
= 417*10E+06 (ft-lb mass) / (h.h.lb force)
h
f
= Nusselt condensing film conductance BTU/(ft
2
.h.F)
k = Thermal conductivity of cooling water BTU/(h.ft
2
.F)
k
f
= Thermal conductivity of condensate film BTU/(h.ft
2
.F)
k
m
= Thermal conductivity of tube material BTU/(h.ft
2
.F)
LMTD = Log mean temperature difference F
MW = Generated power MW
n = Number of passes in compartment
Q = Heat transfer rate to cooling water BTU/h
PF = Performance Factor %
R
a
= Thermal resistance attributed to air ingress F/(BTU/(ft2.h))
R
f
= Thermal resistance due to fouling for tube pairs F/(BTU/(ft2.h))
R
fc
= Increase in thermal resistance due to both fouling and air
ingress F/(BTU/(ft
2
.h))
R
fmean
= Mean fouled thermal resistance of tube pairs F/(BTU/(ft
2
.h))
R
t
= Thermal resistance of cooling water film F/(BTU/(ft
2
.h))
R
w
= Thermal resistance of tube wall F/(BTU/(ft
2
.h))
T
in
= Cooling water inlet temperature F
T
out
= Cooling water outlet temperature F
T
v
= Vapor saturation temperature F
U
clean
= Heat transfer coefficient calculated from clean tube pairs BTU/(ft
2
.h.F)
U
eff
= Overall condenser effective heat transfer coefficient BTU//(ft
2
.h.F)
U
foul
= Heat transfer coefficient calculated from fouled tube pairs BTU/(ft
2
.h.F)
U
HEI
= Heat transfer coefficient based on HEI tube bundle value BTU//(ft
2
.h.F)
U
ref
= Reference heat transfer coefficient based on
sum of clean-tube thermal resistances BTU/(ft
2
.h.F)
v = Cooling water velocity ft/s
w = Mass cooling water flow through compartment water boxes lb/h
T = Temperature gradient across condensate film F
= Latent heat of condensate BTU/lb
= Viscosity of cooling water lb/(h.ft)
f
= Viscosity at condensate film temperature lb/(h.ft)
= Density of cooling water lb/ft
3

f
= Density of condensate film lb/ft
3

4-54
References
1. Code on Steam Condensing Apparatus, ASME PTC 12.2-1983, 1983, publ. ASME, New
York, NY.
2. Performance Test Code on Steam Surface Condensers, ASME PTC.12.2-1998, 1998, pub.
ASME, New York.
3. Tsou, John L., New methods for Analyzing Condenser Performance, Proceedings 1994 EPRI
Heat rate Improvement Conference, Baltimore, MD, May 1994.
4. Kern Donald Q., Process Heat Transfer, 1990 Edition, publ. McGraw Hill, New York.
5. Rabas, T.J. and Cane, D., An Update of Intube Forced Convection Heat Transfer Coefficients of
Water, Desalination, Vol.44, 1983, pp.109-119, publ. Elsevier, Holland.
6. Standards for Steam Surface Condensers, 9th. Edition, 1995, Heat Exchange Institute,
Cleveland, OH.
7. Putman, Richard E. and Karg, Dale C., Monitoring Condenser Cleanliness Factor in Cycling
Plants, Proc. IJPGC, 1999, San Francisco, July 26-29, 1999, publ. ASME, New York.
8. Instrumentation of the On-Line Condenser Fouling Monitor, EPRI Technical Report TR-
109232.
9. Putman, R. E., and G. E. Saxon, Jr. A Newton-Raphson Method for Calculating Condenser
Performance Based on ASME Single Tube Heat Transfer Data, Proceedings, EPRI Heat Rate
Improvement Conference, Dallas, TX, 1996, May 2224, pp. 17-1 to 17-24.
10. Harpster, J.W. An Impact on Plant Performance from Advanced Instrumentation 44
th

Annual ISA POWID Industry Symposium, Orlando, FL (July 7-13, 2001).
11. Harpster, J.W. On Understanding the Behavior of Non-condensables in the Shell Side of
Steam Surface CondensersASME, 2001 International Joint Power Generation Conference,
New Orleans, LA (June 4-7, 2001).
12. Harpster, J.W. Increased Availability from Improved Condenser Design ASME, 2002
Tenth International Conference on Nuclear Engineering, Arlington, VA (April 14-18, 2002).
13. Harpster, J.W. "Reducing Dissolved Oxygen Under Conditions of High Air Ingress"
ASME, 2002 International Joint Power Generation Conference, Phoenix, AZ (June 24-26,
2002).

4-55
GENERATED POWER - MW
C
L
E
A
N
L
I
O
N
E
S
S

F
A
C
T
O
R

-

P
E
R
C
E
N
T
HEI CLEANLINESS FACTOR
ASME PERFORMANCE FACTOR
CHEI = 58.759 + 0.046093 * mw
ASMECF = 43.739 + 0.04745 * MW
WINYAH UNIT #1
FIGURE 1.0
CLEANLINESS FACTORS vs. LOAD

Figure 1
Cleanliness Factors vs. Load, Winyah Unit #1
T
1
T
2f
T
2c
ASME METHOD OF ESTIMATING TUBE FOULING RESISTANCE
FIGURTE 2.0
COOLING WATER
IN
COOLING WATER
OUT
FOULED TUBE
CLEAN TUBE
T
s

Figure 2
ASME Method of Estimating Tube Fouling Resistance

4-56
T
1f
T
s
T
2f
EPRI/BRIDGER METHOD OF ESTIMATING TUBE FOULING RESISTANCE
FIGURTE 3.0
COOLING WATER
IN
COOLING WATER
OUT
FOULED TUBE
PLUGGED TUBE
F
IN-LINE
FLOWMETER

Figure 3
EPRI/Bridger Method of Estimating Tube Fouling Resistance
S
SW
SW
SW
AB
AB AB
SW
SW
SW
Vent Line
Hotwell
Section II
Section I
Section III
Condensate
Tray
Suggested
Tube Pair Areas

Figure 4
Typical Single Shell Single Pass Condenser Configuration

4-57
AIR INLEAKAGE - SCFM
E
Q
U
I
V
A
L
E
N
T

F
O
U
L
I
N
G

R
E
S
I
S
T
A
N
C
E

D
U
E

T
O

A
I
R
D
e
g
.
F
/
B
T
U
/
(
s
q
.
f
t
.
h
)

Figure 5
Typical Relationship Between Air Inleakage and Apparent Increase in Tube Thermal
Resistance

Figure 6
Multi-Sensor Probe and Output Measurement Data (Courtesy of Intek, Inc.)
5-1
5
SESSION 5: FOULING CONTROL
SIDTEC Condenser Cleaning for Cooling Water Systems
R. Jones, GE Betz
J. F. Echols, SIDTEC Services, Inc.
S. D. Jones, GE Betz
R. Post, GE Betz
On-Line Automatic Tube Cleaning System and On-Line Self Flushing Debris Filter
Kaveh Someah
Brackett Green WSA, Inc
Mechanical Tube Cleaning: A Brief Tutorial
George Saxon, Jr.
Conco Systems, Inc.

Session 5: Fouling Control
5-3
SIDTEC Condenser Cleaning for Cooling Water Systems
R. Jones, GE Betz, Trevose, PA
J. F. Echols, SIDTEC Services, Inc., Houston, TX
S. D. Jones, GE Betz, Trevose, PA
R. Post, GE Betz, Trevose, PA

Keywords
condenser, tube cleaning, heat rate, performance, oxidizing biocides
Summary
An alternative to on-line sponge ball cleaning systems was commercialized in 1992. The
alternative product named SIDTEC is a non-intrusive, on-line mechanical tube cleaning system
to maintain steam condenser cleanliness in the absence of oxidizing biocides, gaining over one
hundred unit-years of experience. This paper examines the impact of condenser fouling and a
couple of case histories for plants using the SIDTEC Condenser Maintenance Program.
Introduction
Power plant regenerative steam systems heat sink utilize large volumes of water at relatively
low pressures to condense steam. The main condensers effectiveness is greatly dependent upon
the available surface area for heat transfer and that surfaces ability to conduct heat. A
significant consideration to the heat transfer surfaces ability to conduct heat is the cleanliness of
that surface, commonly referred to as the condenser cleanliness factor. As the condensers
cleanliness factor is degraded, the efficiency of the steam generation cycle is reduced because the
heat sink has become less effective. This results in an additional energy requirement to maintain
the same heat cycle. In extreme situations, the heat cycle will become condenser limited. A
condenser-limited cycle is one in which the output of the system must be reduced because the
condenser is providing excessive backpressure (Figure 1). In this case, increased energy input
could result in the forced shutdown of the system due to insufficient condenser vacuum.
Whether condenser tube fouling is causing a cycle to be less than 100% efficient or it is forcing
the cycle to be condenser limited, the end result is an increased cost of operation. Reduced
efficiency will cause increased fuel costs. Furthermore, a condenser-limited cycle will cause
both an increase in fuel costs and a reduction in power production.
Condenser tube fouling can occur as a result of microbiological growth and/or inorganic fouling.
Inorganic fouling may be present in the form of settled suspended solids, such as mud and silt, or

5-4
scales formed from inorganic salts and/or silica. The most prevalent means of mitigating such
fouling to date is through the use of chemicals; biocides, dispersants, and / or scale inhibitors. In
the presence of a variety of conditions, fouling will occur in the absence of a successful
mitigation program. After fouling has occurred and is no longer acceptable, remedial action is
required. Removal of the foulant can be through either chemical or mechanical means.
Most inorganic scale formation takes place under the cycled conditions of cooling water tower
systems and cooling ponds. Organic fouling, caused by microorganisms, microbial slime,
macro-organisms, and deposition of suspended solids, may be found in most all systems from
cooling towers to large once through condenser cooling systems. In large once through
condenser cooling water systems, the most common and most widely used biocides have been
oxidizers, chlorine chemistry or bromine chemistry. Silt and mud accumulation on the
condenser heat exchanger tubes of these large once through systems have long been problems
with which most have learned to live. In these cases, units are allowed to operate at a reduced
efficiency until such time that an off-line mechanical cleaning could be performed.
Without any outside influences, oxidizing biocides, chlorine and bromine have limitations
associated with their use as effective biocides. These limitations include their deleterious effect
on copper bearing alloys (commonly used condenser heat exchanger metallurgy), toxicity to non-
target organisms at effective use concentrations, and the inability to penetrate and/or remove
microbial slime and inorganic material. Additionally, there is usually an excessive maintenance
requirement to the feed equipment of oxidizing biocides due to corrosion and deposition caused
by the chemicals themselves. In the past these limitations alone have caused users of oxidizing
biocides to seek better solutions to eliminate their dependence on oxidizers.
In more recent years, outside influences such as regulatory agencies and conservation
committees have further increased the need for alternatives to oxidizing biocides. Limitations on
the total pounds and/or concentration of primary pollutants such as oxidizing biocides and
copper have gotten tighter as facility discharge permits are renewed. The expectation is that the
trend will continue to be downward. Since oxidizing biocides tend to increase the corrosion rates
of copper and copper bearing metals, tighter copper discharge limits will certainly also effect the
application of oxidizing biocide programs. Such limitations have forced some oxidizing biocide
programs to be designed with less than effective concentrations, duration, and/or frequencies. In
some cases, dechlorination practices through the use of chemical reducing agents, such as sulfur
dioxide, sodium sulfite, or ammonium bisulfite
1
, must be employed prior to discharge of the
main condenser circulating water to the environment.
Given these limitations, other alternatives to oxidizing biocides have been developed and
employed over the years. One of these alternatives, mechanical-cleaning systems may be
employed either on-line or off-line.
Reliance on off-line mechanical cleaning to maintain condenser cleanliness has the obvious
disadvantage of having to take place off-line. Brushes and scrapers, used on the heat exchanger
surfaces, remove microbes, microbial slime, mud, silt, mineral scale, and any other type of
insulating matter on the heat transfer surfaces. Brushes have proven to have limited
effectiveness depending upon the nature of the deposit while scrapers tend to give the maximum
effectiveness often to a fault, as removal of tube metal is not uncommon. Regardless of the off-

5-5
line cleaning systems efficacy, the fact remains that off-line cleanings means loss production
days. Additionally, performance is allowed to degrade between cleanings.
On-Line Mechanical Tube Cleaning
Prior to the emergence of SIDTEC
, on-line mechanical cleaning was synonymous with the

sponge ball. In these systems, sponge balls are passed through the condenser and collected using
a set of collection grates located within the discharge piping of the Units main condenser. The
limitations of the on-line sponge ball type cleaning systems are as follows:
High initial capital cost for equipment and installation
Extended outage required for installation
Intrusive equipment requiring shut down to perform preventative and corrective maintenance
Collection grates increase main circulating water system pressure, impeding system flow
potentially causing a negative impact on system performance
Possible poor distribution of sponge balls depending upon waterbox configuration and
sponge ball condition
Impingement of sponge balls on collection grates and subsequent passing of the same
through the grates causing cleaner loss with economic and environmental considerations
Installation of the on-line sponge ball mechanical cleaning systems has been wide spread, with
mixed results. The limitations of the sponge ball systems were considered during the
development of SIDTEC with the emphasis on cleaning effectiveness and reliability.
Cleaning effectiveness is directly related to the device used for cleaning. The SIDTEC Rocket
tube cleaner is composed of two components, the body and the cleaning element (see Figure 2).
The body is made of common polymers and provides buoyancy and the proper orientation within
the tube. The buoyancy of the body is a design parameter, dependent upon the condenser inlet
conditions and the requirements for optimum collection efficiency. The cleaning element,
commonly referred to as the skirt, is a flexible disk that makes contact with the full
circumference of the tube's inner diameter. The skirt parameters are optimized depending upon
the nature of the foulant to be removed. This optimization is accomplished by modifying the
skirts diameter, thickness, and rigidity.
The skirt is designed such that the outer edge contacts the inner diameter of the condenser tube.
The contact area creates a squeegee effect at the point of contact while the tube cleaner moves
down the tube. The squeegee effect makes the cleaner closely approximate an off-line scraper
without the potential for metal loss. It is this excellent cleaning quality of the SIDTEC tube
cleaner which allows it to effectively replace oxidizing and other biocides, as well as, eliminate
the need for frequent off-line mechanical cleanings. The SIDTEC tube cleaner effectively
removes microbiological growth. Additionally, the SIDTEC tube cleaner effectively removes
settled suspended solids such as mud and silt. Given time and the correct number of passes,
aggressive type cleaning elements can also remove mineral scales and even silicate-type
deposits.

5-6
After the cleaning element design, installation and operation of the injection /collection system
was evaluated. The design of the SIDTEC cleaning element lends itself to a recovery concept
utilizing skimmer and boom. The development and testing of the cleaning element as well as the
collection technology occurred in the late 1980s.
In 1992, SIDTEC became commercialized, offered as a service, requiring no capital investment
at time of installation. In most cases, all parts of the system were external to the plant and the
plant's conduits. For multiple plants on a common discharge, a single collection system was
utilized for all units. In most cases, penetrations and/or distribution headers, if required, were
installed via hot tapping or during scheduled mini-outages. Because all equipment was external
and accessible, installation and operation had no impact on the existing dynamics of the current
system and/or system design.
CASE HISTORIES
The case histories presented represent SIDTEC On-Line Condenser cleaning programs that
provide excellent cleanliness factors without the use of oxidizing biocides.
TXU Experience
The first commercial installation of SIDTEC was at Texas Utilities (TU) Martin Lake Steam
Electric Station, in East Texas. The station has three each 750 Mw coal-fired supercritical units.
Each unit has an associated condenser with condenser tubes composed of 304 stainless steel.
Main condenser circulating cooling water is supplied to the plant from a fresh water-cooling
pond. For many years oxidizing biocides were used at the plant with limited success. Standard
chlorination practices were consistently unable to maintain an acceptable level of condenser
cleanliness. During the summer months, elevated lake water temperatures and increased system
biofouling had a significant negative impact on the performance of all units at Martin Lake. A
manganese type deposit further complicated the fouling mechanism. The oxidizing biocide,
required to attempt to control the biofilm growth, further exacerbated the manganese deposition.
3

In April of 1992, SIDTEC was installed at the Martin Lake plant while the units remained in
operation. While it is well understood that the condenser cleanliness factor can have a
tremendous impact on a units performance, it is also well understood that other factors at a
steam generating power plant can also have a great impact on unit performance. As inlet
circulating water temperature decreases, the impact of the condensers cleanliness factor on the
units performance declines. A sophisticated plant monitoring station was installed at Martin
Lake to insure that all of the required performance monitoring criteria could be accurately
collected and properly analyzed. Additionally, on-line video borescope inspections were
performed to correlate the data collected with visual observation of the tubes' waterside surfaces.
After on-line cleaning was started, unit performance improved dramatically and visual
inspections verified that cleaner tube surfaces were obtained.
Since 1992, SIDTEC has been the sole method used to maintain condenser cleanliness at TUs
Martin Lake facility. Unit performance has been significantly improved since converting from
oxidizing biocides to SIDTEC. Following the success of the program at Martin Lake, the

5-7
SIDTEC technology is in use at three other TXU coal-fired power plants. At TXU Unit
performance was improved while the requirement for oxidizing biocides were eliminated.
PECO Energy Experience
PECO Energys Peach Bottom Atomic Power Station, PBAPS, is a nuclear generating station
with two each 1200 Mw nuclear powered electric generating units. Main circulating condenser
cooling water is supplied from the Susquehanna River. The condensers on each Unit have
titanium heat exchanger tubes. Tubes constructed of titanium have higher design cleanliness
factor requirements than those constructed of yellow metals and stainless steels. Additionally,
titanium, much like stainless steel, has no natural biocidal effect due to the toxicity of its
corrosion products. Yellow metal corrosion by products do tend to exhibit such a biocidal effect.
Hence, achieving and maintaining cleanliness on titanium heat exchange surfaces is much more
difficult than attaining the same on yellow metal surfaces.
In previous years, PBAPS would perform an off-line condenser cleaning on each condenser
during that unit's corresponding outage. Between outages, sodium hypochlorite, an oxidizing
biocide, was added to the condenser cooling water in an attempt to maintain condenser
cleanliness at an acceptable level. Chlorine was added to each of three condensers, per unit, in
succession. The concentration added to any one condenser was such that the stations NPDES
discharge limit was not exceeded. The Susquehanna River at PBAPS often has high suspended
solids, or mud and silt. Additionally, the river periodically has high concentrations of
manganese. Chlorination of the circulating water oxidizes the manganese and causes it to
deposit on the condenser tubes as a manganese oxide. Initially the deposits are soft and sticky.
Over time, chlorination causes the manganese deposit to form a thin tenacious manganese stain.
Both the initial deposit and the stain are very insulating, as is the deposited mud and silt.
While chlorine applications at Peach Bottom had some success, the condenser cleanliness trend
between outages, twenty-four month period, was absolutely downward. The degrading
cleanliness trend was evident in lost production. During outages, inspections revealed that while
biofilm contributed to the downward performance trend, a large portion of the deposits on the
tubes was inorganic. Much of the deposit consisted of mud, silt, and manganese. Chlorine was
consumed, by the organic constituent of these deposits, making it less effective on any biofilm
mixed in the matrix. Furthermore, chlorination was a major contributor to the manganese
deposition.
In an effort to stop the downward performance trend, Peach Bottom turned to on-line condenser
cleaning. Both SIDTEC and sponge ball systems were investigated by the stations engineering
and chemistry departments. SIDTEC was chosen because it had no impact on outage scope and
it could be installed with the units on-line. Additionally, the capital and engineering cost
required was minimal to that required for three condensers and six conduits with sponge ball
injection and collection equipment. Additionally, consideration was given to the superior
cleaning ability of the SIDTEC rocket tube cleaner. This superior cleaning ability was
demonstrated at a fossil plant on the West Coast of Florida verses plastic plugs during an off-line
demonstration.
3

5-8
SIDTEC was installed at PBAPS initially only on Unit 3. On-line cleaning started with a
modified chlorination practice, which reduce the rate of sodium hypochlorite injection. After the
evaluation period, the chlorination on PBAPS Unit 3 condenser was eliminated and replaced by
weekly on-line cleanings. After successfully demonstrating that improved condenser cleanliness
could be maintained in the absence of chlorination, the decision was made to utilize the existing
SIDTEC system to replace chlorination on PBAPS Unit 2 condensers. In 1997, following the
Unit 2 outage, chlorination was suspended on Unit 2 and SIDTEC replaced the same.
To date, chlorination is no longer utilized for either Units main condenser. Condenser
performance is consistently maintained above that achieved during chlorination. The
improvement is attributed to the ability of on-line cleaning to effectively remove biofilm and silt
and manganese type deposits. Additionally, manganese deposition is no longer exacerbated by
the presence of an oxidizer. Off-line condenser cleanings and the associated cost have also been
removed from Peach Bottoms outage scopes and budgets.
Mississippi Power Co. Experience
Mississippi Powers Plant Daniel is a coal-fired power plant located on the Pascagoula River.
The plant has two each 500 Mw units. The units' were initially designed to utilize once through
cooling on their twin shell, dual pass condensers. The condenser tubes were initially constructed
of admiralty brass and no chemical biofouling treatment program was required nor employed.
The units were designed and installed with a sponge ball cleaning system, however it has been
abandoned-in-place due to high operational and maintenance cost. In late 1995, the Unit 1
condenser at Plant Daniel was replaced with a titanium condenser due to degradation of the
original admiralty brass metallurgy. While the tube metallurgy was changed, no biofouling
treatment provisions were made and no chemical biocides were applied to the system. Initial
testing performed during January of 1996, while the condenser was relatively new and still
extremely clean, indicated that the retrofitted condenser exceeded design performance
specifications for the units turbine.
By May of 1996, condenser performance had significantly degraded and heavy biofouling was
found in the condenser tubes. An off-line condenser cleaning was performed and the units
performance was restored. A team was formed at the plant to evaluate a means of eliminating
this problem.
The possible remedies consisted of either chemical biocides or mechanical cleanings.
Mechanical cleanings were chosen. The mechanical options were off-line cleanings (using
brushes and/or plugs) and SIDTEC. While off line cleanings using plugs had demonstrated that
it could be effective, it had the drawback of taking place off line. During the peak generating
seasons, when biofouling would be most prevalent, this presented a real problem. The plant
would have to lose generation to clean the condensers during times of high demand for power.
Such an option would cause a substantial loss in revenue due to lost generation while the unit
was being cleaned. SIDTEC offered a more viable alternative; a mechanical cleaning that could
effectively take place without removing the unit from service.
Figures 3 and 4 show the results from the first cleaning of the Unit 1 low-pressure (LP) and high-
pressure (HP) condenser over the time period of June 1- July 13, 1998. Unit 1 has 7/8" 18 and

5-9
24 BWG titanium tubes. The data presented is the terminal temperature difference (TTD) and a
calculated cleanliness factor based on the HEI method. The intervals of cleaner injection are
marked on the graphs by horizontal bars. To better display the impact of the on-line cleaning,
load points in a narrow band around the daily full loadwere extracted. This eliminates any
zero samples and data points representing lower loads. There is also a gap in the data between
7/2 and 7/6, due to instrumentation problems over a holiday. The graphs of TTD and %Cl show
the stepwise change in each of these key condenser performance indicators with each
application.
Following the successful application in Unit 1, the first cleaning in Unit 2 occurred between
August 12 -14, 1998. Unit 2 also has a twin shell, multi-pressure condenser with 119 BWG
90/10 CuNi tubes. The on-line, condenser cleaning occurred over two days, circulating Rocket
tube cleaners through the units existing sponge ball injection ports. During the circulation
period, 143,000 passes gave an average of 10.6 passes per tube. Figures 5 and 6 show the
condenser performance improvement for the LP and HP condensers.
SIDTEC is the current means of alleviating biofouling on the Plant Daniel Unit 1 condenser.
Even on the titanium surface, which is more susceptible to biofouling and heat exchange related
problems; SIDTEC is a viable option for biofouling control at Plant Daniel.
Progress Energy Experience
Brunswick is a two-Unit 867 NPR Mw BWR with a once-through cooling water system off the
Cape Fear River in North Carolina. In the spring of 1996, during a scheduled refueling outage
for Unit 2, the plant installed a foul release coating to the inlet circulating water piping in
conjunction with the long term goal of eliminating gaseous chlorine from the site. For the first
several months, condenser performance remained adequate while the sponge ball system was
operational. As is the case in many BWRs, malfunctions and routine maintenance of the ball
system is problematical due to the radiation exposure for personnel in the area around the
condenser. As shown in Figure 7, condenser performance began to degrade throughout the fall
as shown by the terminal temperature difference (TTD) increase from a range of 12 -16 F. to a
range of 37 - 44 F. In December, condenser backpressure was nearing 3.8HgA or two inches
above design, and inlet water temperature was nearing a yearly low.
The CCW system for this plant has four CCW pumps with four conduits to four waterboxes with
an interconnection between the conduits. The maximum CCW flow with four pumps is 1,390
cfs. State restrictions on water withdrawal limit CCW flow to 900 cfs for nine months during the
year. During three summer months, flow is allowed to increase to 1150 and 1210 cfs for the two
units. In December, chlorination was resumed and flow increased using the fourth CCW pump.
This technique improved conditions, reducing condenser backpressure almost 1.5 HgA.
However, with inlet water temperature due to increase through the spring, the plant made plans
to clean the condenser off-line in April to prepare for the summer.
The BetzDearborn team met with plant personnel in January 1997, to discuss using
SIDTEC
Rocket tube cleaners to restore condenser performance in lieu of an off-line

cleaning. Rockets would be injected into the CCW system using the existing sponge ball system
ports and recovered in the discharge using the SIDTEC skimming technology.

5-10
Installation of the recovery equipment began on February 25. Figure 8 shows the general
arrangement of the skimming boom in the canal. Installation of the recovery system was
completed in eight days. The rudimentary system comprised of modified oil spill control boom,
a SIDSKIM tube cleaner recovery unit, a self-priming centrifugal pump, lay-flat discharge hose,
and a de-watering/storage hopper.
During the condenser cleaning, the inventory of tube cleaners would be manually hauled through
security and into the basement to the ball collectors. Each unit has two ball collectors, one each
for the Aand Bside of the condenser. An average shotor load for the ball collector was
3,000 cleaners. Within one to two hours, the entire inventory of 20,000 to 30,000 tube cleaners
could be loaded into the ball collector and shot through the condenser.
The time between injection and the discharge flume was about five minutes. Approximately
one-third of the cleaners were manually recovered in the concrete flume. The remainder would
pass over the weir into the discharge canal to be recovered by the SIDSKIM and pumped back to
the de-water/storage hopper. Generally, one cycle would be completed in the morning and one
in the afternoon.
On March 5, the first tube cleaners were injected to the B-N and B-S side of the condenser.
Figure 9 shows the almost instantaneous response of the condenser to the first load of 3,000.
Referring back to figure 7, the improvement in heat transfer on the Bside was so rapid, the
Aside TTD was reduced 2-3 F. without any cleaners being circulated through that side.
Figure 10 table shows the improvement over the initial three-day circulation period. Gross
generation increased approximately 10 Mw.
Hundreds of tube cleaners were seen with material removed from the tube surface packed
between the cleaning element and the tube cleaner body. Samples of the material were collected
from the tube cleaner bodies and analyzed (See Figure 10).
This analysis is consistent with a silt/biofouling matrix deposit.
The second day of injecting tube cleaners started on the Aside of the condenser. After three
loads, a ball diverter valve became clogged with the remnants of deteriorated sponge balls and
prevented further circulation. After plant personnel were able to clean out the valves, Unit 2 was
completed on April 9.
Unit 1 was known not to be as fouled as Unit 2. Injection of tube cleaners through Unit 1 took
place on March 24-26. Figure 11 shows the TTD and inlet water temperature from January
through April 1997.
During the cleaning, far fewer cleaners were recovered with material remaining on the tube
cleaner. Instead of the black organic matter found in Unit 2, Unit 1s material was more reddish-
orange, indicative of a higher silt content with less organic matter. During the cleaning, the
initial results showed a 2-3 F. reduction in TTD. However, after plotting data over a longer
period of time, the data resolution showed an improvement in the 1 F range.

5-11
In conclusion, demonstrable performance improvement was obtained by cleaning a heavily
fouled condenser. The BetzDearborn team was able to assemble the equipment and install a
functioning on-line tube cleaning system within six weeks of the first meeting, four weeks from a
purchase order. Condenser backpressure was reduced from almost 2.6HgA to less than 1.9
HgA. Net generation increased upwards of 10 Mw.
Conclusion
Oxidizing biocides are the most commonly used biocides in once through cooling applications.
However, oxidizing biocides have both environmental and efficacy concerns, especially when
applied on the more exotic metallurgies being incorporated into today's condenser designs. The
advent of these metallurgies, which are more susceptible to condenser waterside fouling,
accompanied by the lowering of oxidizer concentrations on NPDES permits can place many
plants designed with once through main circulating water systems in a precarious situation.
Additionally, oxidizing biocides can contribute to waterside deposition when applied in waters
that contain manganese. Furthermore, oxidizing biocides have no effect on inorganic muds and
silts, which inevitably exist on condenser tubes as a minor or sometimes major portion of the
waterside deposition. Hence, while oxidizing biocides do provide some efficacy in the
elimination of biofilms in once through systems they are by no means perfect for such
applications.
In the past oxidizing biocides have been the solution of choice because the effects of the
chemicals on the environment are known and the residuals are easily consumed. Available
options were evaluated on a case by case basis as being cost prohibitive; having their own
efficacy concerns, or offers greater environmental concerns. The futures optimal solution must
be to control the fouling to the maximum extent possible, without oxidizing biocides maintaining
maximum plant performance with the minimization or elimination of off-line mechanical
condenser cleaning.
BetzDearborn SIDTEC On-Line Mechanical Condenser Maintenance Program offers a viable
alternative to both biocides and off line cleanings. The program, as demonstrated by the case
histories, provides increased condenser cleanliness over the application of oxidizing biocides. It
also eliminates the requirement for corrective off line mechanical cleanings. By doing this,
SIDTEC provides better condenser performance resulting in an improved unit heat rate and
increased unit production. Improved heat rates translate to reduced fuel costs. Additionally,
SIDTEC impacts budgets by eliminating frequent off-line cleanings and their associated cost.
SIDTEC can be installed with the Unit on-line without impacting production.
References
1. Betz Handbook of Industrial Water Conditioning. Trevose, PA: Betz Laboratories, Inc.,
1991, pp 203.
2. Echols, J.F. and Larry Morris. "A New Approach to On-Line Mechanical Tube Cleaning." in
Condenser Technology Conference Proceedings. J. Tsou, Chairperson. EPRI. St. Petersburg,
Florida. September, 1993.

5-12
3. Echols, J.F. and Jones, R. "Five Years Operating Experience with the SIDTECOn-Line
Mechanical Condenser Tube Cleaning System." in 58th Annual International Water
Conference Proceedings. J. Schubert, Chairperson. Engineer's Society of Western
Pennsylvania. November 1997.

0
20
40
60
80
100
120
140
160
180
200
H
e
a
t

R
a
t
e

I
m
p
a
c
t
Back Pressure
Aux Power
LP Turbine Ef f
Excess Air
HP Turbine Ef f
IP Turbine Ef f
Makeup %
Unburned Carbon
Feedwater Inlet T
Coal Moisture
Flue Gas Temp
Steam Temp
Reheat Temp

Figure 1
Relative Impact of Condenser Fouling on Heat Rate (Btu/kW-Hr)

5-13

Figure 2
SIDTEC Rocket tube cleaner

Figure 3
Unit 1 HP Condenser Performance, June 1 - July 13, 1998
7.1
8.2
9.7
80%
92%
87%
(2.0)
-
2.0
4.0
6.0
8.0
10.0
0
6
/
0
1
/
9
8
0
6
/
0
5
/
9
8
0
6
/
0
8
/
9
8
0
6
/
1
0
/
9
8
0
6
/
1
1
/
9
8
0
6
/
1
6
/
9
8
0
6
/
1
7
/
9
8
0
6
/
2
9
/
9
8
0
7
/
0
1
/
9
8
0
7
/
0
2
/
9
8
0
7
/
0
9
/
9
8
T
e
r
m
i
n
a
l

T
e
m
p
e
r
a
t
u
r
e

D
i
f
f
e
r
e
n
c
e

(
T
T
D
)
70%
80%
90%
100%
110%
120%
130%
140%
C
o
n
d
e
n
s
e
r

C
l
e
a
n
l
i
n
e
s
s

(
%
C
l
)
HP TTD
HP %Cl
Cleanings

5-14

Figure 4
Unit 1 LP Condenser Performance, June 1 - July 13, 1998

Figure 5
Unit 2 HP Condenser Performance, August 12 -14, 1998

3.8
5.6
7.2
113%
156%
128%
(4.0)
(2.0)
-
2.0
4.0
6.0
8.0
0
6
/
0
1
/
9
8
0
6
/
0
5
/
9
8
0
6
/
0
8
/
9
8
0
6
/
1
0
/
9
8
0
6
/
1
1
/
9
8
0
6
/
1
6
/
9
8
0
6
/
1
7
/
9
8
0
6
/
2
9
/
9
8
0
7
/
0
1
/
9
8
0
7
/
0
2
/
9
8
0
7
/
0
9
/
9
8
T
e
r
m
i
n
a
l

T
e
m
p
e
r
a
t
u
r
e

D
i
f
f
e
r
e
n
c
e

(
T
T
D
)
90%
110%
130%
150%
170%
190%
210%
230%
250%
C
o
n
d
e
n
s
e
r

C
l
e
a
n
l
i
n
e
s
s

(
%
C
l
)
LP TTD
LP %Cl
Cleanings
3.8
5.6
7.2
113%
156%
128%
(4.0)
(2.0)
-
2.0
4.0
6.0
8.0
0
6
/
0
1
/
9
8
0
6
/
0
5
/
9
8
0
6
/
0
8
/
9
8
0
6
/
1
0
/
9
8
0
6
/
1
1
/
9
8
0
6
/
1
6
/
9
8
0
6
/
1
7
/
9
8
0
6
/
2
9
/
9
8
0
7
/
0
1
/
9
8
0
7
/
0
2
/
9
8
0
7
/
0
9
/
9
8
T
e
r
m
i
n
a
l

T
e
m
p
e
r
a
t
u
r
e

D
i
f
f
e
r
e
n
c
e

(
T
T
D
)
90%
110%
130%
150%
170%
190%
210%
230%
250%
C
o
n
d
e
n
s
e
r

C
l
e
a
n
l
i
n
e
s
s

(
%
C
l
)
LP TTD
LP %Cl
Cleanings

5-15

Figure 6
Unit 2 LP Condenser Performance, August 12 -14, 1998

-10
0
10
20
30
40
50
1
0
/
2
4
/
9
6
1
1
/
7
/
9
6
1
1
/
2
1
/
9
6
1
2
/
5
/
9
6
1
2
/
1
9
/
9
6
1
/
2
/
9
7
1
/
1
6
/
9
7
1
/
3
0
/
9
7
2
/
1
3
/
9
7
2
/
2
7
/
9
7
3
/
1
3
/
9
7
3
/
2
7
/
9
7
4
/
1
0
/
9
7
4
/
2
4
/
9
7
T
e
r
m
i
n
a
l

T
e
m
p
e
r
a
t
u
r
e

D
i
f
f
e
r
e
n
c
e

(
D
e
g

F
)
40.0
50.0
60.0
70.0
80.0
90.0
100.0
110.0
120.0
130.0
140.0
I
n
l
e
t

W
a
t
e
r

T
e
m
p

(
D
e
g

F
)
A-N
A-S
B-N
B-S
CW T
Clean "B" Boxes
Clean "A" Boxes

Figure 7
Terminal Temperature Difference, Oct 96 - April 97

8.3
10.7
16.0
95%
78%
58%
6.0
8.0
10.0
12.0
14.0
16.0
18.0
8
/
1
2

0
:
3
4
8
/
1
2

7
:
1
4
8
/
1
2

1
3
:
5
4
8
/
1
2

2
0
:
3
4
8
/
1
3

3
:
1
4
8
/
1
3

9
:
5
4
8
/
1
3

1
6
:
3
4
8
/
1
3

2
3
:
1
4
8
/
1
4

5
:
5
4
8
/
1
4

1
2
:
3
4
8
/
1
4

1
9
:
1
4
T
e
r
m
i
n
a
l

T
e
m
p
e
r
a
t
u
r
e

D
i
f
f
e
r
e
n
c
e

(
T
T
D
)
40%
50%
60%
70%
80%
90%
100%
110%
120%
C
o
n
d
e
n
s
e
r

C
l
e
a
n
l
i
n
e
s
s

(
%
C
l
)
LP TTD
LP %CL
Clnr Circ

5-16

Figure 8
Boom arrangement in discharge canal

790
795
800
805
810
10:00
3/3/97
22:00
3/3/97
10:00
3/4/97
22:00
3/4/97
10:00
3/5/97
22:00
3/5/97
10:00
3/6/97
22:00
3/6/97
G
r
o
s
s

G
e
n
e
r
a
t
i
o
n

(
M
w
)
1.50
1.75
2.00
2.25
2.50
2.75
3.00
C
o
n
d
e
n
s
e
r

B
a
c
k
p
r
e
s
s
u
r
e

(
"
H
g
A
)
Mw
BP
Begin Cleaning

Figure 9
Initial response of condenser to on-line tube cleaning

5-17
Performance Improvement

Start Back Pressure Inlet Water Terminal Temperature Difference (F.)
3/5 (HgA) Temp (F.) B-N B-S A-N A-S
08:45 2.56 60.6 33.1 28.4 17.9 19.3
14:45 2.13 61.6 19.4 21.1 15.0 16.8
3/7
12:50 1.90 60.6 15.5 18.3 14.0 12.9

Primary Composition
Loss on ignition 29%
Silicon, SiO
2
22% Sodium Chloride, NaCl 10% Phosphate, P
2
O
5
1%
Iron, Fe
2
O
3
17% Manganese, MnO
2
4% Sulfate, SO
3
1%
Aluminum, Al
2
O
3
12% Magnesium, MgO 3% Calcium, CaO 1%
This analysis is consistent with a silt/biofouling matrix deposit.

Figure 10
Performance Improvement and Deposit Analysis

0
5
10
15
20
25
30
1/24/97 2/7/97 2/21/97 3/7/97 3/21/97 4/4/97 4/18/97
T
e
r
m
i
n
a
l

T
e
m
p
e
r
a
t
u
r
e

D
i
f
f
e
r
e
n
c
e

(
d
e
g

F
)
40
50
60
70
80
90
100
110
I
n
l
e
t

W
a
t
e
r

T
e
m
p
e
r
a
t
u
r
e

(
d
e
g

F
)
A-N
A-S
B-N
B-S
CW T
Clean 3/24 - 3/27
Add 4th CCW Pump

Figure 11
Condenser performance, CP&L Brunswick Unit 1 Jan-April, 1997

5-19
ON-LINE AUTOMATIC TUBE CLEANING SYSTEM AND ON-LINE SELF
FLUSHING DEBRIS FILTER
Kaveh Someah
Brackett Green WSA, Inc.
2969 South Chase Avenue 1335 Regents Park Drive
Milwaukee, WI 53207 Houston TX 77058
Ksomeah@bgwsa.com
414-481-4133

Abstract
The penalty associated with micro and macro fouling within the condenser tubes is significant. It
reduces the power plants efficiency and lowers generation output and availability, while
creating frequent outages and increasing maintenance.
The presence of coarse debris such as marine crustaceans (zebra mussels, blue mussels, shrimp
and fish), wood, rocks, twigs as well as fibrous material such as algae, seaweed, leaves and
plastic bags block the condenser tubes preventing maximum flow. The result is a reduction in
condenser performance and an increase in plant outages and maintenance.
The biological growth as well as fouling and scaling inside the condenser tube also reduces the
overall heat transfer within the condenser resulting in an increased turbine back pressure and
reduction in plant output.
In this paper a typical On-Line Self Flushing Debris Filter and Automatic Tube Cleaning System
for power plant applications will be discussed. Various case studies will be used to demonstrate
the actual plant operating and maintenance cost savings from these systems.
Introduction
The penalties and costs associated with the operation of a fouled condenser are significant. They
include increase in pumping power consumption, fuel consumption, reduced turbine output,
frequent condenser cleaning cost and loss of plant availability.
Macro fouling of condensers with debris such as tiny fish, clams, barnacles, zebra and blue
mussels, jelly fish, leaves, rocks, twigs, seaweed, plastic bags and bottle caps can significantly
reduce the cooling water flow, increase condenser back pressure and results in induced localized
corrosion and erosion of the condenser tubes. Micro foulants such as slim, bacterial matrix,
algae, silt, scale, corrosion products, mud and sludge adhere to the condensers inside tube
reducing the heat transfer rate and turbine output.

5-20
The analysis to estimate the annual power replacement cost or loss of power production for a
given increase in condenser back pressure requires the specific heat rate analysis for a specific
plant. However, Figure 1 can be used with the purpose of establishing an order of magnitude for
such power replacement cost vs. increase in back pressure for a typical coal fired or nuclear
fueled power station.
The cost of manually cleaning the condensers water box and the tubes can also be significant.
Many condensers on the Gulf of Mexico must be manually cleaned as often as every two weeks
during the fouling season. The cost of manual cleaning can approach 500,000 USD per year
including material, labor and lost revenue from reduced power generation (1).

600 MW 1150 MW
COAL FIRED UNIT NUCLEAR FUELED UNIT
CAPACITY FACTOR = 65% CAPACITY FACTOR = 75%
Figure 1
(Energy Replacement Cost vs Increase in Backpressure)
0
1
2
3
4
5
6
7
8
9
10
0 0.5 1.0 1.5 2.0 2.5 3.0
INCREASE IN BACKPRESSURE
ABOVE 1.0" Hga for FOSSIL UNIT
ABOVE 1.5" Hga FOR NUCLEAR UNIT
A
N
N
U
A
L

E
N
E
R
G
Y

R
E
P
L
A
C
E
M
E
N
T

C
O
S
T

(
M
I
L
L
I
O
N
S

O
F

D
O
L
L
A
R
S
)
NUCLEAR UNIT FOSSIL UNIT

5-21
Various methods for debris removal and off-line and on-line cleaning systems are available and
are used by various plants. The off-line cleaning systems often employ the use of hydro blasting
techniques, scrapers, cutters, vibrators, water lances, chemical cleaning, nylon or metallic bristle
brushes, etc. All the mentioned methods are manually applied and require periodic extended and
costly plant shutdowns. Furthermore, since the condenser operates for long periods of time
between each manual cleaning cycle, the fouling initiates and the tubes remain in fouled
condition, reducing the condensers performance and adversely affecting the plants heat rate.
A typical installation for an On-Line Automatic Tube Cleaning System and On-Line Automatic
Self Flushing Debris Filter with respect to the condenser and the cooling water inlet and outlet
flow is shown in Figure 2.

Figure 2
(A Self Flushing Debris Filter & Automatic Tube Cleaning System)
On-Line Self Flushing Debris Filter
Bar screens, trash rakes and travelling band screens is the first line of defense against the
incoming debris with the cooling water intake flow. However debris which passes through or is
carried over by screens as well as debris growth such as zebra or blue muscle and fish within the
cooling water tunnel between the screen and the condenser often accumulate within the
condenser water box or inside the condenser tubes.
Installing an On-Line Automatic Self-Flushing Debris Filter at the condenser inlet nozzle or
within the cooling water inlet pipe can eliminate the debris accumulation within the condenser
(See item 6, Figure 2). Several types of filters are available, including the turbulence, multi
chamber, backwash, hydro cyclone, volute and the pressure relief type. The latest and the most
effective technology used over the last twenty years have been the pressure relief type.

5-22
The pressure relief backwash type filteruses the water from the clean side to relieve, remove
and discharge the accumulated debris from the filter system.
The basic structure for a typical debris filter of this type is shown in figure 3.

Figure 3
(Typical Components of a Pressure Relief Backwash Type Filter)
The filtering screen element is housed within a filter body with the same diameter as the cooling
water supply line (See item 1, Figure 3). The filter screen could be in the shape of a cone, half
circle, dome or a flat screen depending on the manufacturer. The filter screen opening size is
designed to trap the debris and retain it within the screen area only allowing water without any
debris to pass through to the condenser. A typical cone and a half circle type debris filter is
shown in Figure 4.

Figure 4
(A Cone and Half Circle Type Debris Filter)

5-23
The differential pressure across the screen element is continuously monitored (See item 9, Figure
3). Once it reaches the preset value; the rotor drive motor is actuated (See item 6, Figure 3)
while the debris discharge valve is opened (See item 3, Figure 3). The rotor begins to operate at
a low speed while covering a segment (See item 2, 3 & 4, Figure 3) of the screen and the
collected debris. The lower pressure in the debris discharge pipe and rotor housing causes a
reverse flow of a small portion of the filter water through the filter screen element. This flow
reversal results in backwashing and removing the debris from the filter screen and directing it to
the debris discharge pipe that is often connected to the condenser cooling water return line (See
item 5, Figure 3).
Figure 5 demonstrates the complete debris accumulation, cleaning and removal cycle.

Figure 5
(Debris Filter Prior and After a Cleaning Cycle)
As reported by PSI Energy, the plant operates four (4) turbine driven boiler feed pumps for Unit
1 and Unit 2, each unit rated at 500 MW. The main condenser is equipped with backwashing
capability, however cleaning the auxiliary condensers requires shutting down the boiler feed
pump and opening and cleaning the condensers. During such a cleaning cycle the generating
capability of the plant was limited. According to the plant from 1993 to 1996, it was necessary
to perform such cleanings on average of 120 times per year. The annual plant power generation
loss due to auxiliary condenser shut downs from 1993 to 1996 has been over 2,000 MW hr per
year (2).
The Cayuga Station installed the On-Line Automatic Self-Flushing Debris Filter in 1995 and
1996. Since the installation, the condenser tubes have remained clean and free from debris and
the pressure drop through the condenser has remained steady.
The condition of the tube sheets and tubes for two separate water boxes prior and after the
installation of debris filter is shown in Figure 6. The installation of the debris filters at the inlet
to the condenser is shown in Figure 7.

5-24

Figure 6
(Typical Condenser Tube Condition Prior and After Installation of a Debris Filter)

5-25

Figure 7
(Typical Installation of Debris Filters at the Inlet to the Condenser)
On-Line Automatic Tube Cleaning System
The on-line tube cleaning system is used to clean the condenser tubes from micro foulant, scale
and other deposits continuously while the plant remains on line and in its full production.
Slightly oversized special elastomer balls (Figure 8) are periodically or continuously injected
into the condenser cooling water inlet.

Figure 8
(Various Size and Types of Elastomeric Cleaning Balls)
The balls are injected in such a way to provide uniform distribution inside the water box. Since
the balls are slightly larger than the tube ID, they wipe away any deposits which may form on the
tubes. After the balls leave the condenser tubes they are captured by means of a special strainer
that is installed in the cooling water return line (See item 2, Figure 2). The captured balls are
directed toward a collector (See item 4, Figure 2) by means of a small open-recessed impeller

5-26
pump (See item 3, Figure 2). The pump and the collector is used to pump the balls back into the
cooling water supply line creating a closed loop ball circulation.
A major component of this cleaning system is the ball strainer which houses screens that are
mounted on a shaft that can be turned to a backwash position to clean it from any accumulated
debris.
The pressure drop across the strainer screen is continuously monitored by a pressure differential
transmitter. Once the transmitter reaches its set point position, the control panel directs the ball
collector to close and remove the balls from circulation. After the balls are collected, the screen
is automatically opened by means of its motorized actuators and the screens are cleaned against
the incoming cooling water. After the backwashing is completed, the screens close again
automatically and the balls are released into the circulation again. Figure 9 shows the strainer
screen in the closed (normal operating) and open (backwash) position.

Figure 9
(Strainer in Closed (Operating) and Open (Backwash) mode)
A Typical ball re-circulation system and the control panel are shown in Figure 10.

Figure 10
(Typical Ball Recirculation Pumpskid / Collector and Control Panel)

5-27
EPCOR (Edmonton Power) operates two (2) 413 MW coal fired power generating units at the
Genesee Generating Station. The condensers operating at this station experienced massive tube
fouling which greatly effected the turbines back pressure resulting in plant load reduction. The
plant has traditionally dealt with the tube fouling by derating each of the units and by cleaning
one half of the condenser at a time. Both sides can be cleaned within a 24-hour period. The
plant is also able to clean the tubes by injecting a chemical enzyme solution. The chemical
cleaning process also requires that the unit be derated to approximately 170 MWe for 13 hours.
In the summer of 1997 a total of seven (7) derates on both units were taken to chemically clean
the condensers. This amounted to a total of 9900 MW hours deration for each unit (2).
In 1998 the plant installed an On-Line Automatic Tube Cleaning System. The condenser
cleanliness factor for each unit prior to and after the installation of the cleaning system was
reported at 60% and better than 90% respectively. To date, the plant has been able to operate at
the new high cleanliness factor without any need for manual or chemical cleaning of the
condenser. A typical installation for a cleaning system is shown if figure 11.

Figure 11
(Typical Installation of Automatic Tube Cleaning System Including the Strainers, Ball
Recirculation Pump / Collector Skid & Control Panel)

Conclusion
In todays competitive market for the utility industry, the efficiency and the availability of power
generation is a prime concern of the owners.
The use of On-Line Self-Flushing Debris Filters and On-Line Automatic Tube Cleaning Systems
have proven to be the most effective way to eliminate macro and micro fouling within
condensers. The use of such systems will increase the plants heat rate, condenser cleanliness,
overall efficiency and plants availability while reducing tube failure, plant maintenance and
scheduled or un-scheduled outages.

5-28
References
1. Mussalli, Y.G., Bell, R.J. Impagliazzo, A.M., High Reliability Condenser Design Study.
Presented in EPRI Report CS 3200, Palo Alto, Calif.: Electric Power Research Institute, July
1983.
2. Someah, K., Stauber, J., Miller, J., Whitten, D., Peterson, L. On-Line Automatic Tube
Cleaning System & On-Line Self Flushing Debris Filterpresented in Condenser
Technology Conference Palo Alto, Calif.: Electric Power Research Institute, September
1999.

5-29
Mechanical Tube Cleaning: A Brief Tutorial
George Saxon, Jr.
Conco Systems, Inc.

Conco Systems
Mechanical Tube Cleaning
A Brief Tutorial
George Saxon, Jr.
Conco Systems, Inc.
EPRI Condenser Technology Conference
September 11-12, 2002

5-30
Conco Systems
The Tutorial Includes
Introduction to mechanical tube cleaning
Diagnostic techniques
Tube cleaning variables
Deposit characteristics
Various types of mechanical tube cleaners
Case(s)

Conco Systems
Plant Cleaning Preferences
1% Other methods
5% Chemical Cleaning
14% Sponge rubber balls
14% High-pressure water
22% Brushes
44% Metal cleaners
Plant
Preference
Cleaning Method

5-31
Conco Systems
Mechanical Tube Cleaning
Minimizes unit downtime normal crew can clean
5,000 tubes during a 10-hour shift
Effective on all types of deposits:
Fouling deposits
Corrosion products
Physical obstructions
Tube surface roughness
Best cleaning procedure can be developed off-line
through heat transfer testing of fouled tube
samples

Conco Systems
Mechanical Tube Cleaning (cont.)
Deposit sampling capability
Uniform cleaning of each tube
Extend tube life and tube plugging criteria
Retards pitting and improves tube surface
condition
Achieves expected performance improvements
Improves Eddy Current results
Can be performed quickly 10-20 ft./sec.
Safe to use

5-32
Conco Systems
Innovative Cleaner Designs
Hex cleaner
Removed 7,000 lbs. of deposit and increased
capacity by 25 MW saving $4.3 million.
Calcite cleaner
Removed 80 tons of scale and returned
condenser performance to best level in years
saving $2.6 million.

Conco Systems
Hex Cleaner Calcite
Remover
Innovative Tube Cleaners

5-33
Conco Systems
Metal Brush

Conco Systems
U-Tube Cleaner

5-34
Conco Systems
Performance Indicators
Back pressure deviation
Cleanliness factor
Inlet and outlet cooling water temperature
differential
Heat rate
Megawatt output

Conco Systems
Diagnostic Techniques
Deposit Sampling
Heat Transfer Testing
Fouling Monitor
Performance Analysis on the PC
Borescopic Examination

5-35
Conco Systems
Whole-Tube Deposit Samples
Deposit Density
0.1449 1.79
0.4007 4.95
2.0452 25.26
2.9448 36.37
7.5738 93.54
7.7374 95.56
7.9851 98.62
8.9956 111.10
10.2118 126.12
10.5170 129.89
17.5945 217.30
Density
gm./sq.ft.
Dry Weight
gm.

Conco Systems
Results of Elemental Analysis
Not Listed Loss on ignition
Not Listed Elements <0.01%
0.1-1.0 Chlorine
1-5 Calcium
0.1-1.0 Nickel
0.1-1.0 Sulfur
5-10 Silicon
0.1-1.0 Titanium
0.1-1.0 Phosphorus
5-10 Iron
0.1-1.0 Potassium
0.1-1.0 Aluminum
10-20 Manganese
% Element

5-36
Conco Systems
Tube Cleaning Variables
Cooling Water System
Tube Material
Deposit Type(s)
Corrosion
Tube Quantity and Size
Schedule
Tube Cleaning System

Conco Systems
Cooling Water System
Cooling Water Type
fresh water
brackish water
sea water
Cooling System
once through or closed cycle
multiple pass
Water Chemistry
Obstructions and Debris

5-37
Conco Systems
Copper Alloy Tubing
Tubing
Admiralty Brass
Aluminum Brass
Copper-Nickel
Corrosion Characteristics
Crevice corrosion, pitting
Dealloying-dezincification or
denickelification
Erosion-corrosion
Inlet end
Down tube
Ammonia corrosion, stress
cracking corrosion

Conco Systems
Stainless Steel Tubing
Tubing
304
316
AL6XN
Seacure
AL 29-4-C
Corrosion Characteristics
Crevice corrosion, under-
deposit corrosion, pitting
Microbiological influenced
corrosion (MIC)
Chloride pitting

5-38
Conco Systems
Titanium Tubing
Corrosion resistant
Still must be kept clean

Conco Systems
Deposit Types
Particulate Fouling
Crystallization Fouling
Biofouling
Macrofouling
Corrosion Product
Silt and mud
Scale, manganese
and iron
Slime and organisms
Shells, clams,
mussels
CuO
2

5-39
Conco Systems
Ranking Thermal Impact
of Deposit Types
Microbiological
Manganese
Iron
Silicon
Calcium Carbonate
Calcium Phosphate

Conco Systems
Tube Quantity and Size
Quantity
How many
waterboxes?
How many tubes?
Main Section?
Air Removal Section?
Size
Outer Diameter (O.D.)
Gauge (BWG) or (I.D.)
Length
Other considerations
Coatings
Inserts

5-40
Conco Systems
Schedule
Unit Size
Availability
Off-line or reduced load
Time requirement for cleaning

Conco Systems
Tube Condition
Condition Resolution
Deposits 9
Obstructions 9
Corrosion 9
Pitting 9
Surface
Roughness
9

5-41
Conco Systems
The Tube Cleaning System
Performance improvement
Most effective deposit removal
Corrosion protection
Remove obstructions
Clean each and every tube, maintain
consistency throughout
Smooth tube surface
Ease of use

Conco Systems
Various Types of Cleaners
C4S
C3S
Hex
C4SS
H-Brush
XL Brush
Type P
Type CB Cal buster
U-tube Cleaner
All purpose tube cleaner
Hard deposits
Thin tenacious deposits
AL6x and Seacure
Light silt and mud
Light silt and mud
For the lightest deposits
Thick scale
Feedwater heaters

5-42
Conco Systems
Tube Cleaners

Conco Systems
Water Gun
Mobile Pump System
Water Gun and Pump System

5-43
Conco Systems
Tube Cleaner in Action

Conco Systems
Tube Cleaning
Select the most effective
tube cleaner.
Insert the tube cleaners
into each tube.
Utilizing the water gun
and pump system the
cleaners are shot
through the tubes.

5-44
Conco Systems
Inside the Waterbox
Confined Space
Lighting
Planking
Scaffolding

Conco Systems
Post Cleaning
Inspect the condenser
with high intensity light
Rod all obstructed
tubes
Plug designated tubes
Sign-off

5-45
Conco Systems
Cleaning Heavy Deposits From
Heat Exchangers with a HydroDrill

Conco Systems
Expectations
Most effective deposit removal
Immediate improvement in performance
Immediate improvement in heat rate
Reduced outage time
Increased unit availability
Extended tube life

5-46
Conco Systems
Additional Benefits
Economic Benefit
Immediate Return on Investment
Reduced Costs
Recovery of lost megawatts or increased
generation capacity
Fuel savings
Reduction in CO
2
emissions
Extended useful life of the condenser

Conco Systems
Cases

5-47
Conco Systems
Cleaning Results
BACK PRESSURE DEVIATION IMPROVEMENT - NYLON BRUSH vs. METAL CLEANER
FIGURE 11.1
B
A
C
K

P
R
E
S
S
U
R
E

D
E
V
I
A
T
I
O
N

-

p
s
i
a
NYLON BRUSH
METAL CLEANER
NYLON BRUSH
MONTHS

Conco Systems
Cleaning Results
B
A
C
K

P
R
E
S
S
U
R
E

D
E
V
I
A
T
I
O
N

-

in
s
.
H
g
.
MONTHS
HYDROBLAST
CLEANED
MECHANICALLY CLEANED
BACK PRESSURE DEVIATION IMPROVEMENT
FIGURE 11.2
HYDROBLAST CLEANING vs. MECHANICAL CLEANING

5-48
Conco Systems
Cleaning Results
YEAR
B
A
C
K

P
R
E
S
S
U
R
E

-

in
s
.
H
g
.
PARTIAL RETUBE
MECHANICAL CLEANING
DESIGN BACK PRESSURE
PARTIAL RETUBING vs. MECHANICAL CLEANING
FIGURE 11.6
CASE STUDY 4: IMPROVEMENT IN BACK PRESSURE DEVIATION RESULTING FROM

Conco Systems
Potential Savings
from Improving Back Pressure
Pressure Deviation
(inches Hg)
(Excess Back Pressure)
Turbine Rating
(MWatts)
100 400 600 1,000
0.1 $31,250 $125,000 $187,500 $312,500
0.2 62,500 250,000 375,000 625,000
0.3 91,750 375,000 562,500 937,500
0.4 125,000 500,000 750,000 1,250,000
0.5 156,250 625,000 937,500 1,562,500

Source: Bulletin #940-2, INTEK, Inc. 1999 and Operation and
Maintenance of Steam Surface Condensers, Fossil Plant News, Electric
Power Research Institute, Spring 1988

5-49
Conco Systems
Cost of Mechanical Cleaning
625 MW Unit
Description Costs
Labor $19,542
Tube Cleaners 14,276
Equipment 2,860
Expense 3,442
Total $40,120

Note: No Capital Investment is Required

Conco Systems
ROI Evaluation
Cash Flows NPV ROI
System Out In CF
out
-CF
in
NPV/CF
out

Conco $40,120 $562,500 +$522,380 +1,305 %

Assumptions: recover .3 inches Hg
return on 600 MW is approximately the
same as 625 MW

5-50
Conco Systems
Typical Results
Pre and Post Cleaning
Before After Chg.
A Back Pressure Inches Hg. 3.6 2.8 -0.8
B Back Pressure Inches Hg. 4.3 2.9 -1.4
Circ Water
Velocity
Ft/sec 5.7 6.5 +0.8
Net Generation MW 659.4 686.5 +27.1

Conco Systems
Other Cases
Peach Bottom
1,152 MW
River water
55,080 Ti tubes
Manganese
Results
Recover 25 MW
$4.3 Million savings
7,000 lbs. of deposits
Clinton
985 MW
Fresh water, closed system
53,160 304 ss tubes
Calcium carbonate
Results
Recover 20 MW
$2.6 Million savings
80 tons of deposits

5-51
Conco Systems
Key Points
Incorporation of effective tube cleaning practices
is essential in todays competitive market.
Cleaning is also performed for corrosion
protection, increasing availability and reliability.
Increased MW output provides immediate ROI,
increasing profitability.
Select the most effective tube cleaning method
based on your site specific variables.

Conco Systems
Contact Information
George Saxon, Jr.
Vice President, Sales and Marketing
Conco Systems, Inc.
530 Jones Street
Verona, PA 15147 USA
Tel: (412) 828-1166
Fax: (412) 826-8255
E-mail: info@concosystems.com
6-1
6
SESSION 6: ALTERNATIVE COOLING TECHNOLOGY
Wet And Dry Cooling---Cost/Performance Tradeoffs
John S. Maulbetsch, Maulbetsch Consulting

Kent D. Zammit, EPRI
Matthew Layton and Joseph OHagan, California Energy Commission
The Impact of Air Cooled Condensers on Plant Design and Operations
Dirk Jaresch, J & W GmbH
Experience Evaluating Condenser Performance and Tube Fouling with the ASME
Performance Test Code on Steam Surface Condensers PTC 12.2 1998
Dale C. Karg, Santee Cooper
John M. Burns, Burns Engineering Services Inc.
Michael C. Catapano, Powerfect Inc.

Session 6: Alternative Cooling Technology
6-3
Wet And Dry Cooling---Cost/Performance Tradeoffs
John S. Maulbetsch
1
, Kent D. Zammit
2
, Matthew Layton
3
and Joseph OHagan
3

Abstract
The allocation of water resources has become an increasingly contentious issue in many areas.
As a result, power generation facilities are more frequently considering dry cooling for plant
steam condensation. Air-cooled condensers, with attendant water savings have been installed
and operated successfully at many plants in the U. S. and abroad. Many more are planned for
use at plants currently in design or construction.
However, the reduction in water use comes at a price. The capital cost of the equipment is
normally higher than that for comparable wet cooling systems. The plant efficiency is frequently
reduced through the use of dry cooling, as is the plant capacity, particularly during the hottest
days of the year.
This paper presents a methodology for selecting optimized designs for both wet and dry systems
and for comparing the performance and costs of the optimized systems. The cost and
performance differences are shown to be functions of plant and turbine heat rate characteristics,
site meteorology and cooling system design choices. Example comparisons are given for new,
gas-fired combined-cycle plants in the 500 MW size range operating at sites representative of
conditions in California.
A breakdown of the many elements of the capital and operating costs is provided to assist in
understanding the effect of important, but highly site-specific, costs on the general cost
comparisons.
A brief discussion of the environmental trade-offs between wet and dry systems is also given.
Introduction
This paper presents the results of a study to define, explain and document the performance,
economic and environmental tradeoffs between wet and dry cooling systems. While much of the
information may be more widely applicable, the focus of the analysis and the case study
characteristics are specific to conditions in California.

1
Maulbetsch Consulting, Menlo Park, California 94025
2
EPRI, Palo Alto, California 94304
3
California Energy Commission, Sacramento, California 95814

6-4
Approach
The systems considered are closed-cycle wet cooling using mechanical draft wet cooling towers
and surface condensers and direct dry systems using mechanical draft air-cooled condensers.
The approach was a case study method where comparisons of the two systems were carried out
at four sites characteristic of the range of conditions in California.
The chosen sites were a high desert site characterized by conditions at Blythe, California, a
northern mountain site characterized by conditions at Burney, California, a Central Valley site
characterized by conditions at McKittrick, California and a Bay Area/Delta Region site
characterized by conditions at Pittsburg, California.
Plant/steam conditions at each site are as follows:
Plant type: gas-fired combined cycle; 500 MW in 2 x 1 configuration
Steam flow: 1,000,000 lb/hr
Steam quality: 95%
Back pressure: 2.5 in. Hga (T
sat
= 109F)
Water Use and Conservation
The impetus for dry cooling is the increasing interest in water supply and conservation in
California and across the country. Electric power generation utilizes water in many ways and in
varying amounts depending on the type of generating plant and the type of cooling system
employed. Table 1 summarizes some nominal values of water use for various plant types with
alternative choices of systems for steam cycle cooling.
Table 1
Water Requirements for Power Generation (in Gallons per MWh of Plant Output)

Plant Type
Steam
Condensing
Auxiliary Cooling and
Hotel Load
Total
Stand-alone steam plant 720

30 750
Simple-cycle gas turbine 0 150 150
Combined-cycle plant
(2/3 CT + 1/3 steam)
240
(1/3 x 720)
110
(2/3 x 150 + 1/3 x 30)
350
Combined-cycle plant
with dry cooling
0 110 110
Stand-alone steam plant
with dry cooling
0 30 30

6-5
Other Environmental Issues
A comparison of environmental impacts associated with wet and dry cooling systems was made
in nine areas, including water use, waste management, including water discharge, hazardous
materials, air emissions, noise, visual resources, public health, biological resources and
agriculture and soils.
In many of these areas, the effects are proportional to the water use and the associated
withdrawal, evaporation, blowdown, drift, and plume production. This is the case for impacts on
waste management, water discharge, hazardous materials, public health, and agriculture and
soils. In these cases, the use of dry cooling eliminates the effects entirely. The other areas are
discussed further below.
Air Emissions
Dry systems eliminate effects associated with drift and volatile components stripped from the
water in wet towers. However, the use of dry systems imposes penalties on plant efficiency and
capacity, which requires that additional fuel be burned at the plant or elsewhere on the system to
produce the same net power generation. This leads to an increase in the emissions associated
with the combustion process. While this is a second-order effect on a state- or system-wide basis,
it can be a measurable effect in the vicinity of the plant and should be considered on a case-by-
case basis.
Noise
The noise from wet systems comes from both the water falling through the tower fill and the fans
and air motion. In the case of dry systems, the water noise is obviously eliminated, but the fan
and air noise may be increased since the quantity of air moved through the system is greater.
The importance of this effect is site specific, but, in locations where noise abatement is
necessary, the use of special low-noise fans or sound walls may be required. For so-called
ultralow- noise fansthe increase in cost is estimated at approximately 10%. Similar noise
abatement may be required for the wet towers in some cases.
Visual Resources
Dry systems effectively eliminate the occurrence of visual plumes. However, they require
physically larger, taller structures than wet cooling and, as such, can be a more obvious element
at a plant site. The importance of this feature is also site specific.
Biological Impacts
The primary focus in this area is on entrainment and impingement losses regulated under Section
316(b) of the Clean Water Act. Dry systems obviously reduce this impact in proportion to the
reduction in total plant water use, estimated in Table 1 as approximately 70%.

6-6
A postulated effect of dry cooling is so-called atmospheric entrainment,where insects and
even small birds may be entrained with the inlet air stream. To date, no research or information
is available on the importance of this issue
Performance and Cost
The reduction in the amount of water used and other related environmental benefits of dry
cooling are technically achievable. However, the benefits come at a cost. In essentially all
situations, the use of water as the cooling medium is the cheapest way to provide cooling at
power plants. Furthermore, wet cooling always results in higher annual plant output and in more
efficient plant performance during most of the year. The relevant questions are how much more
expensive and how great is the effect on plant output and efficiency.
The performance and cost issues are inextricably related. The initial choice of a larger, higher
capacity and, hence, more expensive cooling system will result in higher plant capacity and more
efficient operation for the life of the plant. This is true for all cooling systems, wet as well as dry.
In general, a proper cost comparison among alternative cooling systems must be made between
optimized selections at a particular site. An optimized system would normally be defined as one
that minimizes the sum of all costs---initial capital cost, operating energy cost, efficiency penalty
cost, and capacity penalty cost---for the life of the plant. The magnitudes of these cost elements
are dependent on many variables. Some of the most important include the operating
characteristics of the plant generating components, the meteorology at the plant site, the present
and projected costs of fuel, the present and projected price of power and the projected demand
profiles for the plant. The choice of the optimum or preferred design also depends on the relative
importance assigned to present vs. future costs, which depends strongly on the economic
objectives and business plans of the plant owner. Therefore, the optimum systemor the
system of choicemight range from a system of lowest first cost to one of lowest total lifetime
cost projected for a 30-year or longer lifetime.
To illustrate the methodology for determining the various cost elements and for selecting an
optimumsystem based on a range of criteria, case studies were conducted at four sites. In each
case the cooling system was sized to condense the turbine exhaust steam from the steam portion
of a nominal 500-MW gas-fired combined-cycle plant typical of what is currently being
proposed and built in California.
The cost information and comparisons are based on vendor-supplied data for new, 500 MW gas-
fired, combined-cycle plants using fresh water makeup.
Wet System Costs
Table 2 lists the major cost elements of a wet cooling system.

6-7
Surface Condenser
Budget price estimates were obtained on steam surface condensers for a turbine exhaust flow of
1,000,000 lb/hr @ 5% moisture, a heat load of 980 x 10
6
Btu/hr at a turbine back pressure of 2.5
in. Hga (T
cond
= 108.7F). The tube material is 304 stainless steel; 1OD/20 gauge. The
operating points are a terminal temp. difference (TTD) of 5 F (if available), a range of 20F
and a cold water temperature of 7090 F.
Table 2
Capital Cost Elements for Wet Cooling System Equipment for New Plants
Element Comment Cost
Wet cooling tower Erected tower including structure,
fans, circulating pumps, fill, drift
eliminators, etc.
Strongly dependent on materials,
assumed Douglas Fir; typically 35
to 45% of system cost
Installation/erection Included in base price --
Surface steam
condenser
Major cost element (see previous
subsection)
Typical range of $5 to $12/kWe;
approx. 35 to 45% of system cost
Tower basin Including typical site preparation Significant cost item; function of
tower size; estimated at $25/ft of
basin perimeter plus $10/sq ft of
basin area; typically 3 to 6% of
system cost
Electricals and controls Fan/pump motor wiring and controls,
etc.
Important cost item; estimated at
$25,000 per cell
Circulating water
system
Pumps, piping, valves, etc. Can be significant cost item;
dependent on site layout; assumed
at 5% of total installed cost
Water supply/intake
structure
Highly site dependent; minor if source
is nearby; major, if water supply is far
from site or at much lower elevation
Estimated at 1 to 2% total installed
cost
Water
treatment/blowdown
discharge
Usually minor; may be significant if in
zero-discharge region where
evaporation ponds or brine
concentrators may be required
Estimated at 1% total installed cost
Auxiliary cooling Typically 5% additional heat load Estimated at 5% additional cost
Additional elements Typically minor and site dependent Not included in case study
estimates of comparisons
Site preparation/
access provision
Winter operation;
freeze protection
Low-noise fans
Painting
Fire and lightning
protection
Acceptance testing
Highly site dependent; likely minor; not likely to be affected significantly
by system choice
Location dependent

Significant cost if required; more important for dry systems than for wet
Typically minor costs


6-8
All cases resulted in two- pass tubeside designs, divided water box, carbon steel shell, tube
support plates, tube sheets, air cooling shrouds, and water boxes. Tube side velocities ranged
from 7.5 to 10 ft/sec and the number of tubes from 9000 to 16,000. The prices included an air
ejector air removal package. The prices were quoted ex-works,meaning ready for shipping on
the suppliers loading dock. Shipping, unloading, assembly, and testing are extra to be provided
by others.
For a total cost estimate (including installation), a 20% adjustment was chosen. Table 3 displays
the costs, normalized against area and turbine output (in kWe), along with the log mean
temperature difference (T
ln mean
) for each case. Figure 1 displays the costs in $/kWe plotted
against T
ln mean
showing a smooth relationship that is used for scaling condenser costs for the cost
comparisons.
Table 3
Surface Steam CondensersNormalized Budget Prices
Cold Water
Temp. ( F)
T
ln mean
( F)
Base Price Installed Price
(+ 20%)
Price/Area
($/sq ft)
Price/Output
($/kWe)
70 27.5 $ 790,000 $ 948,000 15.6 5.58
75 22.2 $ 990,000 $ 1,190,000 15.7 7.00
80 16.7 $ 1,240,000 $ 1,490,000 15.4 8.76
85 12.0 $ 1,630,000 $ 1,960,000 13.2 11.5
90 10.4 $ 1,800,000 $ 2,160,000 13.2 12.7

Condenser Installed Costs
($/kWe) = 91.404(LMTD)
-0.8364
0
5
10
15
0 5 10 15 20 25 30
Log Mean Temp. Difference, deg. F
$
/
k
W
e

Figure 1

6-9
Cooling Towers
The base system chosen to represent recirculating wet cooling is the mechanical draft, counter-
flow wet cooling tower. Budget prices were obtained for wet cooling towers for each of the four
sites.
Two criteria were used for each site:
A low first costcase in which the capital cost of the tower was minimized at the expense of
additional fan power; and
An evaluated costcase in which the sum of the capital cost and the cost of power evaluated
over the assumed 30-year life of the tower was minimized.
2,500,000
2,700,000
2,900,000
3,100,000
3,300,000
3,500,000
3,700,000
3,900,000
4,100,000
4,300,000
0 5 10 15 20 25
Approach, deg F
S
y
s
t
e
m

C
a
p
i
t
a
l

C
o
s
t
,

$
Desert Site Mountain Site Valley Site Bay Area Site

Figure 2
Wet Cooling System Capital Cost vs. Approach for Minimum Evaluated Cost Design (for
New 500-MWe Facilities with 170-MWe Steam Cycle)
The cost of the power required to operate the pumps and fans of the cooling system is borne
continuously for the life of the plant. In order to put these continuing future costs on a common
basis with the initial capital costs, they have been converted into an evaluated cost per kW.
The values chosen for evaluating the cooling system power costs are an energy cost of
$60/MWh, a 6.7% discount rate, a 3% escalation, a 50% tax rate, and a 30-year plant life. These
result in an evaluated power cost of $3625/kW. The total evaluated costs as a function of
approach are shown in Figure 3.

6-10
Table 4
Site-to-Site Cost EstimatesWet Cooling Tower and Surface Condenser for New 500-MWe
Combined-Cycle Plants with 170-MWe Steam Cycle
Low First Cost Design
System Capital
Cost
2,924,000 2,710,000 2,820,000 2,680,000
Total BHP (in hp) 1723 1851 1794 1505
-fans (in hp) 1234 1498 1441 1198
-pumps (in hp) 489 353 353 307
Minimum Evaluated Cost Design
System Capital
Cost
3,331,000 3,118,000 3,405,000 2,960,000
Total BHP (in hp) 964 978 1030 987
-fans (in hp) 377 645 713 680
-pumps (in hp) 587 333 317 307

5,000,000
5,500,000
6,000,000
6,500,000
7,000,000
7,500,000
0 5 10 15 20 25
Approach, deg F
T
o
t
a
l

E
v
a
l
u
a
t
e
d

C
o
s
t
,

$

Figure 3
Wet Cooling System Total Evaluated Cost vs. Approach for Minimum Evaluated Cost
Design (for New 500-MWe Facilities with 170-MWe Steam Cycle)

6-11
Dry Cooling
The base system for dry cooling is a direct system with a mechanical draft air-cooled condenser
(ACC). The major items of the capital cost of the system are displayed in Table 4.
Budget prices for a dry cooling tower and associated heat transfer and flow components were
obtained from several major vendors for each of the four sites.
Dry System Cost Analysis
For the dry system cost estimates, vendors were free to select the design point of their own
choosing, based presumably on their engineering and commercial experience with the
appropriate tradeoff between initial cost and performance penalties. The design ambient
temperature chosen for a particular site differed significantly from estimate to estimate, leading
to towers of significantly different size and cost for each location. However, the estimates
showed good consistency across the range of unit sizes (expressed as ITD; defined as T
cond
-
T
amb
). Figure 4 gives the range of capital cost vs. ITD.
10,000,000
15,000,000
20,000,000
25,000,000
30,000,000
35,000,000
40,000,000
45,000,000
50,000,000
10 15 20 25 30 35 40 45 50 55 60
ITD, deg F
C
a
p
i
t
a
l

C
o
s
t

Figure 4
Capital Cost vs. ITD for Air-Cooled Condenser (for New 500-MWe Facilities with 170-MWe
Steam Cycle)
Power Requirements
Figure 5 displays the power requirements at the motor terminals for the range of design ITD's.
The values were normally provided as power required at the motor terminals. In those cases
where the required power was specified as fan shaft power, a motor efficiency of 95% was
assumed.

6-12
Table 5
Capital Cost Elements for Dry Cooling System Equipment for New Facilities
Element Comment Cost
Air-cooled condenser Strongly dependent on choice
of design point expressed as
ITD (T
cond
-T
amb
)
Ranges from $100 to
$250/kWe
Installation/erection Significant cost item; quoted in
different ways
Ranged from $175,000 to
$225,000 per cell; $200,000
used in comparisons
Steam duct support; column
foundations
Installation dependent Estimated for 10
6
lb/hr unit at
$120,000 to $160,000;
$150,000 used for costs and
comparisons
Electricals and controls

Fan/pump motor wiring and
controls, etc.
Important cost item; estimates
ranged from at $20,000 to
$35,000 per cell; used 5.5% of
installed base cost
Auxiliary cooling Typically 5% additional heat
load; typically handled with
separate unit (usually wet) but
occasionally as extra cells on
ACC
Estimated at 7.5% additional
cost without specifying choice
of auxiliary unit
Cleaning system for finned
tube surfaces
Minor but required in most
locations
Estimated at $150,000
Low-noise fans Included in base costs (far-
field sound pressure levels of
~65dBa at 400 feet)
---
Additional elements Typically minor and site
dependent
Not included in case study
estimates of comparisons
- Water supply/intake
structure
Minor (but not zero) for dry systems
- Water treatment/blowdown
discharge
Minor (but not zero) for dry systems
- Site preparation/access
provision
Highly site dependent and likely minor; not affected significantly
by system choice
- Finish paint; fire/lightning
protection
- Winter operation; freeze
protection
Location dependent and relevant to both wet and dry systems;
typically 2 to 4% of total installed cost; not included for
California estimates

6-13
2,000
4,000
6,000
8,000
10,000
12,000
14,000
10 20 30 40 50 60
ITD, deg F
F
a
n

P
o
w
e
r
,

H
P

Figure 5
Fan Power vs. ITD
The capital costs and the evaluated cost of power sum to the total evaluated cost shown in Figure
6.
-
10,000,000
20,000,000
30,000,000
40,000,000
50,000,000
60,000,000
70,000,000
80,000,000
90,000,000
10 20 30 40 50 60
ITD, deg F
T
o
t
a
l

E
v
a
l
u
a
t
e
d

C
o
s
t
,

$
Total Evaluated Cost Capital Cost Evaluated Power Cost

Figure 6
Total Evaluated Costs vs. ITD

6-14
Table 6
Case Study Cost EstimatesAir-Cooled Condenser ((for New 500-MWe Facilities with 170-
MWe Steam Cycle)
Desert Site
ITD = 37
Mountain Site
ITD = 44
Valley Site
ITD = 44
Bay Area Site
ITD = 55
Capital Cost ($) 30,300,000 25,500,000 25,500,000 20,400,000
Fan BHP (in hp) 5300 4770 4570 3560
Total Evaluated
Cost ($)
44,700,000 38,400,000 37,900,000 30,000,000
Determination of Penalty Costs
A cooling system is designed to maintain a certain back pressure for a given heat load at a given
ambient temperature, normally selected well below the maximum temperature expected at the
site during the hottest periods of the year. Therefore, during periods in which the ambient
temperature exceeds the design temperature, the back pressure will be higher than design
resulting in a higher plant heat rate.
Additionally, if the ambient temperature goes too high, the steam flow must then be reduced to
avoid the risk of exceeding the allowable turbine back-pressure limit. This can result in a
significant reduction in the power output from the steam cycle.
Performance curves for the ACC, shown in Figure 7, display the back pressure that can be
achieved at a given ambient temperature for the design heat load.
1
2
3
4
5
6
7
8
9
10
11
55 65 75 85 95 105 115
Tamb, F
B
a
c
k
p
r
e
s
s
u
r
e
,

i
n

H
g
a
ITD=44 ITD=55 ITD=34 ITD=37 ITD=21

Figure 7
ACC PerformanceBack Pressure vs. T
amb
(for New 500-MWe Facilities with 170-MWe
Steam Cycle)

6-15
From the annual temperature-duration profile at the site, the number of hours per year that the
system would operate at or above a given back pressure can be determined. The corresponding
lost output is displayed in Figure 8 below.
y = 2075.5x - 5486.9
0
5,000
10,000
15,000
20,000
25,000
0 2 4 6 8 10 12 14 16
Turbine Back Pressure, in Hga
L
o
s
t

O
u
t
p
u
t
,

k
W

Figure 8
Lost Output vs. Turbine Back Pressure
Figure 9 displays the sum of the total evaluated cost and the penalty costs.
0
50,000,000
100,000,000
150,000,000
200,000,000
250,000,000
300,000,000
20 30 40 50 60
Design ITD, deg F
C
o
s
t
,

$
$50/MWh $60/MWh $100/MWh $250/MWh $500/MWh

Figure 9
Total Evaluated Cost and PenaltiesSteam Side Only

6-16
Extension of Penalty Evaluation Procedure to Other Sites
The results of applying this penalty evaluation procedure to other sites differ only because of
differing site meteorology.
Additional considerations include:
Operating Power Requirements
The energy required to operate cooling system fans and pumps is energy that must be generated
but that cannot be exported or sold by the plant. The cost of this reduced output is incurred for
the life of the plant. It is normally expressed as the present worth of a kilowatt that could produce
energy at a projected price per MWh for the assumed plant life. This value can be traded off
against the capital cost of the cooling system (which can normally be reduced through the
expenditure of higher amounts of fan and pump power).
For wet systems in this analysis, the present worth of the power was approximately equal to the
capital cost of the cooling system for an optimized system and approximately 160% of the
capital cost for a low first costsystem.
For dry systems, the operating power for a given heat load is much greaterby a factor of 4 to
6than that for an optimized wet system.
Operations and Maintenance
Comparisons of O&M requirements were based not on detailed analyses or surveys but rather on
information obtained during interviews and discussions with staff at operating plants equipped
with dry or hybrid systems, all of whom had had previous experience at other plants using wet
cooling towers. In each case, the water-conserving system was described as basically trouble free
and easy to operate. No additional staff was required, nor was any unusual amount of staff time
allocated to the cooling system beyond occasional scheduled cleaning of the finned tube
surfacesand this could be performed with the unit on line.
Therefore, no additional costs were assigned to dry or wet/dry systems for extra O&M
requirements.
Summary
Initial capital costs for wet and dry cooling systems that could provide suitable cooling for the
four case study sites vary greatly, ranging from $2.7 to $4.1 million for wet systems and from
$18 to $47 million for dry systems. Figure 10 displays the initial capital costs for wet and dry
cooling systems of both lowest first costand optimizeddesigns for each of the four sites.

6-17
DESERT
MOUNTAIN
VALLEY
BAY AREA
Wet---Low First Cost
Wet---Optimized
Dry---Low First Cost
Dry---Optimized
0
5
10
15
20
25
30
35
C
a
p
t
i
a
l

C
o
s
t
,

m
i
l
l
i
o
n
s

$
Case Study Site

Figure 10
Case Study Cost Summary

6-19
The Impact of Air Cooled Condensers on Plant Design and Operations

Richard E. Putman Dirk Jaresch
Technical Director Managing Director
Conco Consulting Corp. J & W GmbH
Verona, PA 15147 Bochum, Germany

Abstract
Air-cooled condensers were first introduced into the U.S. power industry in the early 1970s, but
only during the last decade has the number of installations greatly increased, largely in response
to the growing attention being paid to environmental concerns. The rising importance of this
rather different technology for the condensing and recovery of exhaust steam calls for a broader
understanding of the associated design and application principles involved, as well as of the
performance monitoring techniques and cleaning methods that have to be applied.
This paper identifies the basic configurations of air-cooled condensers used in the power industry
together with their advantages and disadvantages when compared to those exhibited by
traditional steam surface condensers. The several factors that affect the performance of air-
cooled condensers are described in detail, especially the consequences that result from the
fouling of the finned-tubes. Measuring the performance of air-cooled condensers is, clearly, an
important task and some methods are identified, alternative ways of presenting the data also
being included.
To rectify the degradations in performance that result from external tube fouling, a number of
cleaning procedures are described. Included among these are details of a new automated cleaning
technology that has been successfully applied, and some significant performance improvements
that have resulted from the use of this technique are documented.
Keywords: Air-cooled Condensers, Dry Cooling, Performance Monitoring, Cleaning,
Maintenance
Introduction
One of the important features of the Rankine cycle, on which all fossil, nuclear and combined
cycle power plants are based, is the condensation of the vapor exhausted from the LP turbine
stage, the condensate being recycled back through the system. Not only does this reduce the
amount of makeup water that has to be supplied and treated but the heat contained in the
condensate is also recovered. There are three major condensation systems that are employed:

6-20
Steam surface condensers with once-through water cooling paths
Steam surface condensers with closed cooling water systems that include either forced or
natural draft cooling towers
Systems with air cooled condensers
Once-through Water Cooling
Once-through condenser water cooling systems take water from a lake, river or sea inlet and
discharge it back into the source. While being relatively inexpensive, they can also increase the
likelihood of the tubes silting up or fouling. In some cases, environmentalists have real concerns
about an excessive heating up of the cooling water source during the summer months, so causing
damage to the surrounding ecosystem.
Systems with Cooling Towers
Although more expensive in installed cost, closed systems with cooling towers have the
advantage that cooling water chemistry is better able to be controlled, so reducing loss of
condenser performance due to fouling. However, todays plant designer also has to take aesthetic
considerations into account. The concrete hyperbolic cooling tower structures that are a part of
so many plants may not be allowed in a pristine rural area. The plume from cooling towers may
also introduce problems, especially if there is the possibility that the accompanying fog will,
without warning, make an adjacent highway dangerous to drive.
Systems with Air Cooled Condensers
Over the past 30 years there has been a growing and competing demand for water for both
domestic and industrial use and this has brought an increased interest in the use of air as a
cooling medium in place of water. In the utility industry, the earliest applications for the air-
cooled condensing of exhaust steam were modified air-cooled heat exchangers similar to those
already in use by the process industries. Eventually, air-cooled condensers designed for the
utility industry evolved into a configuration that recognized the special needs of condensing a
large volume of low pressure vapor as well as the removal of non-condensibles.
Even so, air-cooled heat exchangers are still sometimes used to offset the thermal pollution
resulting from the cooling water discharge from steam surface condensers. For instance, the
Plant Branch station of Georgia Power is passing a parallel stream of river or lake water through
a set of air-cooled heat exchangers in which the temperature of the water stream is reduced by as
much as 20 Deg.F. (Blankinship, 2001)
(1)

The air-cooled condenser that evolved to meet these needs, while it has some disadvantages, has
been able to provide a solution to some otherwise quite intractable design problems. In addition
to providing a practical solution to some of the problems identified above for both once-through
and cooling tower systems, air-cooled condensers make it possible to build a power plant in
locations without adequate cooling water resources, often the case when a power plant is to be

6-21
built at the mouth of a coal mine. Thus, in spite of certain cost and other disadvantages, air-
cooled condensers offer an important alternative to more conventional cooling water systems
and, since 1968, some 65 installations are already operating in various parts of the world, many
on combined cycle plants.
Air Cooled Condensers
Figure 1.0 shows a general view of an air-cooled condenser typically of the A-frame design.
Berger(1992)
(2)
indicates that not only does this design facilitate condensate draining and
collection but it also ensures that there are no dead zones in the heat transfer surface, that there is
a high operating stability during load transients, while it also eliminates freezing even with
ambient temperatures as low as 58 Deg.F (-50 Deg.C).
The vapor inlet header constitutes the apex of the A. A large-diameter and comparatively
lengthy pipe connects this header to the exhaust from the low pressure stage of the turbine and its
large volume makes this inlet subsystem prone to air inleakage as well as requiring a longer time
to evacuate during plant startup.
At the bottom of the A-frame are two outlet headers, each connected to the inlet header by banks
of finned tubes as shown in Figure 2.0. Most of the panels on an air-cooled condenser are of this
parallel flow type, in which both the condensing vapor and condensate flow together down the
insides of the tubes. Piping also connects the two outlet headers together, allowing vapor to pass
from one side to the other as well as the condensate to be collected.
Examination of Figure 1. shows dearator (or dephlegmator) panels located towards the center of
the banks of tubes. A cross-section through these panels is provided in Figure 3. and it will be
seen that the vapor and condensate are in a counter flow arrangement, the vapor rising up into the
tube banks from the outlet headers while the condensate flows back down to these headers so
that it can be collected and withdrawn. Meanwhile, the upper ends of the tubes in these sections
are connected to their own headers, which are also provided with steam jet air ejectors for the
removal of non-condensibles.
The finned tubes are necessary because of the low thermal conductivity, low density and low
heat capacity of air. The larger surface area required to obtain a given heat removal rate, the area
increasing with the design ambient air temperature, also means that the footprint of air-cooled
condensers is larger than their water cooled equivalents. This is indicated in Figure 4, the air-
cooled condenser provided to Black Hills Power and Light plant in Wyoming, which also shows
the type of support structure required. One other problem is the noise created by the large
number of fans, which may introduce its own environmental problem. Finally, an air-cooled
condenser is more expensive that its water-cooled equivalent. However, in spite of these negative
attributes, the air-cooled condenser has been adopted as the equipment of choice for at least 65
installations as the only way to provide new sources of power in difficult locations.

6-22
The Environmental Protection Agency and Dry Cooling
In 2000, the U.S. Environmental Protection Agency
(3)
conducted a comparative study of the
environmental impacts of wet vs. dry cooling. Their conclusion was that the energy consumption
per lb. condensate was higher for dry cooling than for wet cooling and that the atmospheric
emissions associated with that energy consumption was also higher. The energy penalty also
increases with the ambient air temperature. These disadvantages are offset by the cooling water
intake flow being reduced by 99% over that required by a once-through system; or 4-7% over a
closed cooling water system. They also noted that dry cooling eliminates visual plumes, fog,
mineral drift and water treatment and waste disposal issues. However, their conclusion was that,
dry cooling does not represent the best technology available (BTA)for minimizing
environmental impact.
Much of the E.P.As concern is that the high costs and energy penalty of dry cooling systems
may remove the incentive for replacing older coal-fired plants with more efficient and
environmentally favorable new combined-cycle facilities, the latter presumably equipped with
wet-cooling systems. Their general concern is understood but should not prevent plants being
built using air cooled condensers where there is no practical alternative.
Fouling Tendencies of Air Cooled Condenser

The external surfaces of the finned tubes on air-cooled condensers are very prone to fouling from
pollen, dust, insects, leaves, plastic bags, bird carcasses, etc. Not only is the air flow affected but
also the heat transfer coefficient, the deterioration in performance increasing unit operating costs.
In severe cases, fouling can also limit the power generation capacity of the turbogenerator.
To improve the heat removal capacity of an air-cooled condenser under conditions of high
ambient air temperature, operators will sometimes spray water on the heat exchanger to reduce
surface temperature. Unfortunately, depending on the quality of water used, this sometimes leads
to new scale formation on the tube fins and, again, reduces the heat transfer rate if the deposits
are allowed to accumulate.
Performance Calculation Principles
Several standards exist for calculating the performance of air-cooled heat exchangers and it
would seem to be approaching an exact science. Among these standards are ASME PTC.30
(4)
,
API Standard 661
(5)
and the Standards for Air-cooled Heat Exchangers published by the Air-
cooled Heat Exchanger Manufacturers Association
(6)
.
However, there are no standards at this time for calculating the performance of air-cooled
condensers, nor can the standards for air-cooled heat exchangers be applied. One main difference
is that, while air-cooled heat exchangers with their fans are built as discrete units, the fans
provided with air-cooled condensers are not uniquely associated with a corresponding bank of
tubes. Thus, when a fan is switched off or its speed reduced, not only is the air flow to all tube
banks in the condenser reduced but the distribution of the air among the tube banks can also

6-23
change. Some air-cooled condensers are also equipped with programmable logic systems that
adjust fan speeds, vanes, etc. automatically to ensure that subcooling of the condensate does not
occur but this, again, affects the distribution of the air. Fouling of the tube surfaces can also
affect air distribution. Finally, local meteorological and ambient conditions have their own
effects on the performance of air-cooled condensers.
Kroger
(7)
outlines in detail a method for calculating the performance of air-cooled condensers
from first principles, based on an extensive knowledge of the condenser design data.
Unfortunately, this data is not readily available and the calculations are complicated. The
following is, therefore, an attempt to gage the performance of air-cooled condensers empirically,
using a selected set of operating conditions as the frame of reference.
It is suggested that there be two references cases, both assuming that the turbogenerator is
running at full load. Case(a) would be with all air-cooled condenser fans running at full speed
and Case(b), also with the turbine at full load but with the fans running at half speed. The reason
for having two reference cases is that, in cold weather, it may not be desirable to run the fans at
full speed. The condenser should be calibrated when clean for both of these cases, using at least
the set of instrumentation indicated in Figure 5.0, the values being averaged across all banks.
Among the criteria to be captured for the reference cases, against which subsequent performance
can be compared, are:

Pressure of air at inlet to tube banks - P
ai

Pressure of air leaving tube banks - P
ao

Pressure drop across the tube banks P
tb
= P
ai
- P
ao
(1)
Corresponding air inlet temperature - T
ai

Corresponding air outlet temperature - T
ao

Vapor saturation temperature - T
s

Condenser backpressure - P
s

Pressure of air at fan inlet - P
fi

Pressure of air at fan outlet - P
fo

Pressure drop across the fans - P
fan
= P
fo
- P
fi
(2)
Fan speed - N
rpm

Condenser duty Q = W
cond
*(H
vap
- H
liq
) (3)
Ambient air temperature - T
amb

Effective modified heat transfer coefficient: U
mod
= A
eff
* U
eff
(4)

Using this combined function means that the effective surface area of the tube banks does not
need to be known. Assuming that the log mean temperature difference can be calculated from:
LMTD = (T
ao
T
ai
) / logn((T
s
- T
ai
) / (T
s
T
ao
)) (5)
Then U
mod
= A
eff
* U
eff
= Q / LMTD (6)
Meanwhile, the operating data can be presented in several ways:

6-24
One curve that is often available is shown in Figure 6.0, in which the condenser duty is
plotted against inlet dry bulb temperature for various values of condenser backpressure. The
curve in Figure 6.0 is used when all fans are running at full speed and a similar but different
curve is usually available for fans running at half speed. These are in fact condenser capacity
curves and can be calibrated against measured conditions when the unit is first started up and
while the finned-tube banks in the condenser are still clean. Putman
(8)
has shown how,
subsequently, current condenser duty can be calculated from present backpressure and
turbogenerator load, using the low pressure stage expansion lines included in the thermal kit
data provided by the manufacturer of the turbogenerator. This may be compared with the
condenser duty calculated in accordance with equation (3) above.
Another form of data presentation is shown in Figure 7.0, in which condenser backpressure is
plotted against the percent of design air flow and for various values of the ambient air
temperature. The air flow can be estimated from fan characteristic curves using the pressure
at the inlet to the tube banks mounted in the A-frame. The actual backpressure can then be
compared with that expected at 100% air flow for the current ambient air temperature. The
avoidable condenser loss corresponding to this deviation in backpressure can be estimated
using, again, the expansion lines included in the thermal kit data for the low pressure stage of
the turbogenerator.
Of course, the performance of an air-cooled condenser can become degraded not only by the
external fouling of the finned tubes but also by any internal fouling from the condensate (e.g.
ammonia corrosion) or by air ingress into the condensing vapor. Harpster
(9)
has suggested a way
of distinguishing between the effects of fouling and air ingress, using instrumentation applied to
the air removal system, injecting known amounts of air or nitrogen into that part of the system
operating under vacuum and noting the change in effective heat transfer. The air removal system
instrumentation can subsequently be used to estimate the contribution of air ingress to the change
in the effective heat transfer coefficient.
Cleaning Techniques for Air-Cooled Condensers
(10)
The three principal methods for cleaning the external surfaces of air-cooled condensers are as
follows:
Fire hose
High pressure handlance
Automated cleaning machine
Fire Hose
While the volume of water consumed is high, a fire hose offers only a low washing effect
because of the low pressure involved. The galvanized surfaces of the tubes and fins are not
damaged by this method. Unfortunately, in order to perform cleaning the plant must be taken out
of service and scaffolding erected. The process may also be time and labor intensive depending
on unit design and accessibility.

6-25
It has also been found that use of the fire hose only leads to small performance improvements
even if the surfaces seem to be optically clean. The reason is that only a portion of the fouling
material is washed off while the remainder is pressed between the fin tubes and can not be
washed out by this method. Furthermore, once compressed, the fouling material not only hinders
heat transfer but also obstructs air flow.
High Pressure Handlance
The high pressure handlance method offers low water consumption and a high water pressure.
Unfortunately, the latter can cause the galvanized surfaces to become damaged or even cause the
fins to be snapped off. Again, the plant must be taken out of service and scaffolding erected in
order that cleaning can be performed. Unit accessibility will affect cleaning productivity.
As with the use of a fire hose, this procedure only leads to small performance improvements and,
once the fouling material has been compressed, it hinders heat transfer and obstructs air flow.
Automated Cleaning Machine
The automated cleaning machine, an example of which is shown in Figure 8.0, uses a significant
volume of water; but at a pressure that, while allowing for effective surface cleaning, avoids
damaging galvanized surfaces and fins. The main components of the system include a nozzle
beam, a tracking system, and a control panel. The water contains no additives. The nozzle beam
is optimally matched to the tube bundle geometry, with a constant jet angle. Optimizing the
geometry of the nozzle beam involves determining the proper nozzle distance to the surface, the
jet energy and the selection of the appropriate nozzle design. Variations in nozzle beams are
shown in Figures 9.0 and 10.0. The constant jet angle also ensures that there is no damage to or
snapping off of tube fins, regardless of the material from which they are fabricated. Furthermore,
the carriage on which the nozzle beam is mounted moves at a constant speed and so allows the
fouling to be removed effectively and uniformly across the heat exchange elements of the
condenser. Because the fouling material is removed, air flow is no longer obstructed.
An important advantage of the automated cleaning method is that cleaning can be performed
during operation while the unit is still on-line. Further, there is no need for scaffolding and labor
requirements are minimized. The automated cleaning system can be applied in three principal
forms:
a. Permanently installed system complete with PLC controls, one system being supplied for
each side of the condenser as previously shown in Figure 8.0.
b. Semi-automatic system in which only the guide rails are permanently installed, the nozzle
beam carriage being moved from section to section as the cleaning progresses as shown in
Figure 11.0.
c. Portable service unit, together with a portable nozzle beam carriage and control unit. The
cleaning service is performed in-house or by a qualified service provider as shown in Figure
12.0.

6-26
Performance Improvements from Cleaning
Data from several power plants equipped with air-cooled condensers show that, after cleaning to
remove external fouling, it was possible to operate the unit with the fans running at half-speed
rather than full-speed. The lower auxiliary power consumption resulted in a reduction in
operating costs.
In another plant, condenser cleaning resulted in the generated power rising from 15 MW to 18
MW.
To clean an air-cooled condenser installed in a 400 MW power plant located in the United
Kingdom, a semi-automatic cleaning system was used. An analysis of the heat rate deviation
curve for this unit showed that a 1 in.Hg improvement in turbine back pressure was equivalent to
savings of $188.00/h accompanied by an increase in generation capacity of 4 MW.
Turbine back pressure before cleaning = 3.40 in.Hg.
Turbine back pressure after cleaning = 2.62 in.Hg.
Back pressure reduction = 0.78 in Hg.
Savings at a 75% load factor = 0.78 * 188.00 * 7 * 24 * 0.75 = $18,476/week
The data was taken at an ambient temperature of 59 Deg.F and it was found that the air flow
before cleaning was 78% of its design flow rate.
Conclusion
In the power industry, the reduced availability of water as the cooling medium for the
condensation of exhaust steam, combined with an increased emphasis on environmental
considerations, often makes the selection of an air-cooled condenser a viable alternative to the
traditional steam surface condenser. Although their capacity is sometimes limited by ambient
conditions, their selection can avoid a number of other problems, and often facilitate the
acceptance of the proposed plant by the permitting authorities. Further, because the external
surfaces of the finned tubes on the air-cooled condenser are prone to fouling, an effective
cleaning system is required. One such system has been demonstrated. Finally, there is a need to
develop new standards for acceptance test procedures and for calculating their performance
under a variety of operating conditions.
References
1. Blankinship, Steve, (2001), Georgia Powers New Cooling Tower Design Reduces
Environmental Impact, Power Engineering, September 2001, p. 11.
2. Berger, Norbert, Dry Cooling Systems for Large Power Stations, publ. GEA Energietechnik
Gmbh, 1992

6-27
3. EPA Rule 316(b) New Facility Rule, Chapters 3: Energy Penalties and Chapter 4: Dry
Cooling, publ. EPA (2000)
4. ASME PTC.30, (1991) Air Cooled Heat Exchangers, publ. American Society of Mechanical
Engineers, New York, NY
5. API Standard 661, Air-Cooled Heat Exchangers for General Refinery Service, publ.
American Petroleum Institute, Washington, D.C.
6. Standard for Air-Cooled Heat Exchangers, (1986), publ. Air-Cooled Heat Exchanger
Manufacturers Association, New York.
7. Krger, D.G., (1998), Air-Cooled Heat Exchangers and Cooling Towers, publ. Begell House,
New York
8. Putman, Richard E. (2001), Steam Surface Condensers: Basic Principles, Performance
Monitoring and Maintenance, publ. ASME Press, New York, NY.
9. Harpster, J.W. (2001), On Understanding the Behavior of Non-condensables in the Shell
Side Steam Surface Condensers, Proceedings of IJPGC-2001, New Orleans.
10. Mller-Steinhagen, H., (2000), Mechanical Cleaning with High Pressure Water of the
External Surfaces of Air Cooled Heat Exchangers, Heat Exchanger Fouling: Mitigation
and Cleaning Technologies, publ. PUBLICO Publcations, Essen/Germany, pp.76-87.

6-28

Figure 1
General View of Air-cooled Condenser
TURBOGENERATOR
EXHAUST VAPOR
PARALLEL
FLOW
CONDENSER
CONDENSATE RETURN
VAPOR FLOW
FAN
MOTOR
AIR
FLOW
CONDENSATE COLLECTOR
AIR-COOLED CONDENSER
FIGURE 2.0
PARALLEL
FLOW
CONDENSER
AIR
FLOW
PARALLEL FLOW SECTIONS
INLET
HEADER
VAPOR
CONDENSATE

Figure 2
Air-Cooled Condenser, Parallel Flow Sections

6-29
TURBOGENERATOR
EXHAUST VAPOR
STEAM JET
AIR EJECTOR
CONDENSATE RETURN
VAPOR FLOW
FAN
MOTOR
AIR
FLOW
CONDENSATE COLLECTOR
FIGURE 3.0
DEAERATOR SECTION
INLET
HEADER
AIR
FLOW
STEAM JET
AIR EJECTOR
COUNTER
FLOW
CONDENSER
(DEPHLEGMATOR)
(OR DEAERATOR)
VAPOR
CONDEN-
SATE

Figure 3
Air-Cooled Condenser, Deaerator Section

Figure 4
Air Cooled Condenser in Black Hills Plant

6-30
EXHAUST VAPOR
FAN
MOTOR
FIGURE 5.0
INLET
HEADER
T
T
T
P
P
ao
ai
fo
fi
amb
p
ao
p
ai
INSTRUMENTTAION
T
s

Figure 5
Air Cooled Condenser, Instrumentation
INLET DRY BULB TEMP. - Deg.F
C
O
N
D
E
N
S
E
R

I
N
P
U
T

E
N
E
R
G
Y

-

M
B
T
U
/
h
CONDENSER CHARACTERISTICS
ALL FANS AT FULL SPEED
FIGURE 6.0
15 in.Hg
13 in.Hg
11 in.Hg
9 in.Hg]
7 in.Hg
6 in.Hg
5 in.Hg
4 in.Hg
3 in.Hg
2 in.Hg

Figure 6
Condenser Characteristics, All Fans at Full Speed

6-31
CONDENSER BACKPRESSURE CHARACTERISTICS
FIGURE 7.0
DESIGN AIR FLOW - PERCENT
T
U
R
B
I
N
E

B
A
C
K
P
F
R
E
S
S
U
R
E

-

i n
.
H
g
AMBIENT
TEMP.
81.5
77.0
72.5
68.0
63.5
59.0
54.5
50.0
45.5
41.0
36.5
Deg.F

Figure 7
Condenser Backpressure Characteristics

Figure 8
Automated Cleaning Machine

6-32

Figure 9
Variation in Nozzle Beam

Figure 10
Variation in Nozzle Beam

6-33

Figure 11
Semi-Automated System

Figure 12
Portable System

6-35

EXPERIENCE EVALUATING CONDENSER PERFORMANCE AND TUBE
FOULING WITH THE ASME PERFORMANCE TEST CODE ON STEAM
SURFACE CONDENSERS PTC 12.2 -1998
Dale C. Karg
Santee Cooper
Moncks Corner, SC
John M. Burns
Topsfield, MA
Michael C. Catapano
Powerfect Inc.
Livingston, NJ

Abstract
The Performance Test Code on Steam Surface Condensers underwent a major revision in 1998
with the release of ASME PTC 12.2 1998, Performance Test Code on Steam Surface
Condensers. One of the improvements over the previous code revisions was the development
of a new method for determining or estimating tube bundle fouling. Instrumentation has been
developed to apply the new methods and was used in a test program on a recently retubed
condenser. This paper describes the testing program initiated by Santee Cooper at the Winyah
Generating Station where a successful attempt to utilize the guidelines established by the new
Code were carried out.
The objectives of our testing program were to:
Evaluate the data collected from the instrumentation developed to determine tube bundle
fouling.
Propose methods for utilizing the new instrumentation to evaluate tube bundle fouling
overtime.
Compare the results of the test to design based on the Heat Exchange Institute Standards for
Steam Surface Condensers (HEI)[2].
The initial results of the program demonstrate a good correlation between the actual versus
projected condenser performance and tube bundle fouling by employing the methodology
outlined in ASME PTC 12.2 1998. Santee Coppers retubed condenser provided an excellent
opportunity for the application of the new test code since the condition of the replacement heat

6-36
transfer surface was known as virtually clean. The purpose of the testing was solely for
investigative and information purposes as there was no performance acceptancecriteria placed
on the retubing project.

Introduction
The determination of the degree to which waterside tube fouling and heat transfer resistances
influence condenser performance has long been a challenge to conducting accurate condenser
testing. For meaningful results during acceptance testing, the adverse effects of fouling on
condenser performance (pressure) needs to be accounted for. As well, during routine operation,
quantifying the incremental reduction in power output and higher heat rate caused by tube
fouling would allow better informed decisions as to the resultant cost of fouling versus the cost
justification for scheduling condenser cleanings. The 1998 revision to ASME PTC 12.2 [1] now
provides practical procedures to quantify tube bundle fouling and determine the heat transfer
resistances.
The retubing of a condenser in the spring of 2000 at Santee Coopers Winyah Generating Station
in Georgetown, South Carolina provided an opportunity to apply the methods specified in ASME
PTC 12.2 1998. The results of the testing would be compared to the expected performance of
the new tubing based on performance rating methods such as HEI.
As recommended by the test code, it would be necessary to instrument the individual water
temperature of several pairs of condenser tubes to determine the tube fouling resistance. The
instrumentation device was not a readily available product. Powerfect Inc. offered to design and
provide prototype fixtures (tube fouling monitors) based on their experience with patented tube
testing and plugging systems. The requirements of this method are in section 4.3.12, Tube
Bundle Fouling (Cleanliness)of the Code. Burns Engineering Services provided technical
support on optimizing the location of the tube fouling monitors and utilization of the test code.
Condenser Retubing
The original condenser was supplied with 18 BWG Admiralty tubes and 90-10 Cu-Ni in the air
cooler and periphery areas. The condenser cooling water source is from a cooling pond that
receives fresh water for makeup. After 22 years of service the condenser had more than 10% of
its tubes plugged. Condenser tube leaks were a concern to maintaining good boiler water
chemistry. The original condenser showed evidence of at least three failure modes. A large
number of admiralty tubes failed due to steam side ammonia attack below the air removal
section. Other failures were attributed to significant pitting on the tube side, likely due to under
deposit corrosion. Still others were possibly a result of sulfide attack.
To provide better corrosion protection from ammonia attack and avoid significant modifications
to the condenser tube support structure, 90-10 Cu-Ni tube material was selected for retubing the
condenser. Analysis of the performance impact of material changes from 18 BWG Admiralty to
either 18 or 20 BWG 90-10 Cu-Ni was completed. The analysis to determine the change in
design condenser pressure followed the calculation procedures from the condenser

6-37
manufacturers manual, which is based on HEI [2] methods. The original design information is
as follows:
Original Tubes:
Admiralty 18 BWG Main Bank
Quantity 15,088
9010 Cu-Ni 18 BWG Air Cooler & Periphery
Quantity 1,400
Cooling Surface 151,000 Sq. Ft.
Effective Length 35-0
Original Design:
Steam Condensed 1.361 x 10
6
lb./hr.
Heat Load to Circ. Water 1.309 x 10
9
Btu/hr
Temp Inlet Circ. Water 79 F
Temp Outlet Circ. Water 101.77 F
Circ. Water Flow (Max) 115,000 GPM
Tubeside Velocity (Max) 7 FPS
Flow Passes 2
Cleanliness Factor 85%
Absolute Pressure 2.40 in. Hg.
The design condenser pressure was expected to increase 0.10 in. HgA to 2.50 in. HgA if the
tubing was changed from Admiralty to 90-10 Cu-Ni using the original 18 BWG gauge. A
PEPSE thermal heat balance model indicated that the increase in design condenser pressure
would decrease generation by 729 kW and increase heat rate by 18.7 Btu/kWh on this 300,000
kW unit.
With the material selected, a comparison was made between 18 and 20 BWG tubing to determine
the overall expected improvement in performance with better heat transfer provided by the
thinner wall 20 BWG tubing. As mentioned above, the calculated design pressure for 18 BWG
is 2.50 in. HgA, but it decreased to 2.49 in. HgA for the 20 BWG tubing. The small difference
of 0.01 in. HgA in condenser pressure was estimated to decrease heat rate by 2 Btu/kWh and
increase output by 74 kW. There was a significant material cost savings associated with the
thinner 0.035 in. 20 BWG tubing versus the 0.049 in. 18 BWG tubing if it could be utilized
without significant modification to the tube supports.
Santee Cooper contracted Burns Engineering Services to conduct a vibration analysis for the
existing tube support spacing and use of the 20 BWG tubing. As a conservative measure to
reduce vibration potential, the study recommended that the upper tubes be staked in-between
support plates. The staking provided a reasonable cost solution to take advantage of the lower

6-38
tube cost and slight performance improvement of the 20 BWG tubing. The decision was made to
retube with the 20 BWG 90-0 Cu-Ni tubing with 18 BWG 70-30 Cu-Ni tubing in the air removal
and impingement areas.
Performance and Tube Bundle Fouling Test Methodology
A test program following the methods outlined in ASME PTC 12.2 1998 was initiated and
conducted to:
Gain experience with the new test code
Evaluate newly developed instrumentation to measure tube bundle fouling
Propose methods of monitoring tube bundle fouling over time
The test results enabled a comparison of actual versus expected performance for the new
tubing material and wall thickness.
Tube bundle fouling resistance, as per the test code, is determined by measuring the temperature
rise of adjacent pairs of tubes distributed throughout the bundle. One tube of each pair is left in
the normally fouled condition of the condenser, while the other is thoroughly cleaned or replaced
with a new tube. If the condenser is new or retubed, the cleanreference tubes should be
plugged for the period of time between the test and while the condenser is put back in service.
As the reference tubes representing the overall fouled condition of the condenser begin to foul
the fouling resistance of these tubes will increase. When the clean tubes are unplugged just prior
to the test the fouling resistance can be measured as the difference in temperature rise between
clean tubes and those representing the overall fouled condition of the tube bundle. The period of
time will depend on the aggressiveness of the waterside fouling.
The code suggests a pair of tube fouling monitors be installed for each 2,000 tubes but not to
exceed 16 pairs for the condenser. In our case with 15,088 tubes, 8 pairs were installed. During
installation additional pairs should be considered to provide redundancy in case any of the
sensors fail. Depending on the aggressiveness of fouling, access to the condenser waterboxes,
importance of the test and integrity of the installation, consideration to adding as many as 50
percent additional fouling monitors should be considered. Since the installation was a two-pass
condenser application, the inlet temperature of the second pass had to be measured as well.
Fouling resistance is calculated for each pair of tubes and then averaged to represent the fouling
resistance of the condenser per equation 5.1.11 of the code (see the Test Computation Section for
the equation). The data required by the test code:
Common inlet temperature
Condenser pressure
Overall flow
Discharge temperature of each tube pair (in addition to the inlet temperatures to the second
pass for this 2-pass condenser.)

6-39
Tube Bundle Fouling Instrumentation

The main objective was to provide an anchoring fixture that could accurately position the
temperature sensing leads to the selected, subject tubes, without disturbing the cooling water
flow. Care was used to minimize any adverse effects to the overall effort of getting accurate
temperature data.
The tube fouling monitors securely hold the temperature sensor at the outlet of the selected tubes
or inlet (required for two-pass condensers). The instrument wiring from each sensor is run
carefully between tube ligaments and non-mechanically fastened to the tubesheet in several
locations. The wiring exits each waterbox at a single location through a full port ball valve and
is connected to a data logger. Although not the method utilized for our test, a typical sketch from
the test code (Figure G.1) is shown as a method of securing tube fouling monitor sensors:
`

The test code provides the methodology for locating the monitors. With assistance from Burns
Engineering Services, the monitors were located in the approximate centroids of equal tube
sectors and not within the first three periphery rows. The test tubes were also selected to be near
the penetrations that provided the exit for the instrumentation wiring to pass through the
waterboxes.
Test Procedure
The test was conducted with instrumentation traceable to the National Institute of Standards and
Technology (NIST). All pressure and temperature instrumentation was calibrated prior to the

6-40
test. The Code requires that the circulating water discharge temperature include at least one
sensor every 1.5 square feet. This condition was met with 5 instruments of various depths for
each of the 2 discharge lines. The inlet and outlet temperatures were measured with precision,
calibrated RTDs. Condenser pressures were measured with four basket tips and precision
calibrated pressure transmitters. The unit was operated within six percent of design full load
during the test.
With the exception of not being able to measure the circulating water flow, the PTC 12.2 1998
test procedure was followed. While we recognize that the circulating water flow measurement is
a significantly important test parameter it could not be measured during the test. We installed
Pitot tube taps as far downstream of the discharge waterbox as possible but the readings were
very unstable. Readings taken with an ultrasonic meter were also very unstable. We believe
there was too much turbulence in this area.
To estimate the flow, the total dynamic head (TDH) was measured and compared to the flow
versus head system curve. The TDH intersected the flow curve very close to the design flow.
The circulating water flow was therefore assumed to be at design for the purposes of our
demonstration. This method of flow estimation would be acceptable for performance monitoring
but not accurate enough for acceptance testing.
Test Computations
The test code provides detailed calculation procedures in Section 5.0 with an example in
appendix D. All of the following equation numbers refer to the ASME PTC 12.2 1998 test
code.
From the test data the test heat load (Q
+
) is computed from equation 5.1.1 as follows:
Q
+
= w
+
* c
+
p
* (T
+
2
- T
+
1
)

Where:
+
= measured or calculated from test data
Q
+
= condenser heat load (Btu/hr)
w
+
= flow rate (lb/hr)
c
+
p
= specific heat (Btu/lbF)
T
+
2
= outlet temperature (F)
T
+
1
= inlet temperature (F)
From the test data the overall test heat transfer coefficient is calculated using equation 5.1.2:

U
+
= Q
+
/ (A
o
* LMTD
+
)

6-41
Where:
+
U
+
= overall test heat transfer coefficient (Btu/hr- ft
2
- F)

Q
+
= condenser heat load (Btu/hr)
A
o
= effective outside heat transfer surface area (ft
2
)
LMTD
+
=

Log Mean Temperature Differential (F)
The heat transfer relationship between the overall heat transfer coefficient and the sum of the
resistances is then utilized to determine the apparent shell side resistance from the test data. The
fouling resistances are determined from the tube fouling monitors (R
+
f
equation 5.1.11). The
tube metal resistance (R
+
m
equation 5.1.4)

and the tubeside resistance (R
+
t
equation 5.1.6) are also
calculated.
The fouling resistance is calculated from a comparison of the temperature rise from the tube
fouling monitors (R
+
f
equation 5.1.11-simplified version of the equation) for each pair of tubes is:
R
+
f
= [A
o
/ (nw
+
c
+
p
)] * [(1/ ln ((T
+
s
-

T
+
1f
) / (T
+
s
-

T
+
2f
)) -
(1/ ln ((T
+
s
-

T
+
1c
) / (T
+
s
-

T
+
2c
))]

Where:
+

c
= clean reference tube

f
= fouled reference tube
R
+
f
=

resistance due to fouling (hr- ft
2
-F/Btu)
A
o
2
)
n = number of condenser tubes
w
+
= flow rate (lb/hr)
c
+
p
T
+
s
= condenser steam saturation temperature (F)
T
+
2
T
+
1
The individual fouling resistances (R
+
f
) for each pair of fouling monitors are then averaged to
determine the overall test tube bundle fouling. The individual values of each fouling resistance
(R
+
f
) and raw temperature data should be reviewed to determine if any of the values appear to be

6-42
outliers. The review can entail a simple visual comparison of the resistance data or an estimate
of the standard deviation of the fouling resistance values can be developed. In our case, this
analysis indicated a strong possibility that one of the first pass outlet temperatures should have
been labeled as a second pass inlet temperature.
With the fouling resistance (R
+
f
), tube metal resistance (R
+
m
) and tubeside resistance (R
+
t
)

calculated, the shell side resistance can then be determined from the difference of the sum of the
resistances and inverse of the test overall heat transfer coefficient using equation 5.1.12:
R
+
s
= 1/U
+
-

R
+
m
- (d
o
/d
i
) - R
+
f
The test resistance values are then used in the acceptance test evaluation to allow a comparison
of the test pressure adjusted to design reference conditions. For meaningful test results, the test
pressure is adjusted to design reference parameters. This is completed by adjusting test
calculated heat transfer resistances for the differences between the test and design conditions.
The adjusted resistances are used to calculate an adjusted overall heat transfer coefficient and the
test adjusted steam pressure at the design reference conditions. The design condition is usually
targeted in a condenser acceptance test or a monitoring program to have an unambiguous
reference performance from which all deviations are measured. Naturally other references could
be employed but performance related costs, turbine backpressure, opportunity costs, or
incremental generation have often been evaluated only specifically for the design condition.
There is no adjustment to the tubewall metal resistance (R
+
m
) as it is a constant, dependent only
on tube material and wall thickness.
The design tubeside thermal resistance (R*
t
) is calculated at the design conditions using equation
5.2.3:
R*
t
= 0.0451 [*
0.373
/ (k*
0.538
* *
0.835
*c
+
p
0.482
)] * d
i
0.165
/v*
0.835

Where: R*
t
= design tubeside thermal resistance (hr- ft
2
-F/Btu)
* = value derived from design reference
= viscosity (lb/hr-ft)
k = thermal conductivity (Btu/hr- ft

- F)
= density (lb/ft
3
)
c
+
p
d
i
= inner tube diameter (in.)
v* = average tube velocity (ft/sec)

6-43
The design tube fouling resistance is calculated from equation 5.2.6:
R*
f
= (1 c*
f
) / U*

Where: R*
f
= tube fouling resistance adjusted to design (hr- ft
2
-F/Btu)
c*
f
= design cleanliness factor (%)
U* = test heat trans. coefficient derived from design (Btu/hr- ft
2
- F)
Determination of the adjusted shell side condensing heat transfer resistance in reality is very
complex but the design condition can be simplistically projected from the test result by the ratio
of the change in the shell side Nusselt Number. Section 5.2.4 of the test code should be referred
to for the calculation of the shell side resistance corrected to the design reference conditions and
for the purpose of brevity, is not reproduced here.
With all the individual condenser heat transfer resistances defined, the overall heat transfer
coefficient (U
o
) adjusted to the design reference conditions is calculated from equation 5.2.15 as
follows:
U
o
= 1 /

[R*
m
+ R*
t
(d
o
/d
i
) + R*
f
+ R
o
s
]

Where: U
o
= test heat transfer coefficient corrected to design (Btu/hr- ft
2
- F)

o
= test value corrected to design reference conditions
R*
m
= tube material resistance (hr- ft
2
-F/Btu)
R*
t
= tube waterside resistance (hr- ft
2
-F/Btu)
d
o
= outside tube diameter (in.)
d
I
= inner tube diameter (in.)
R*
f
= fouling resistance derived from design (hr- ft
2
-F/Btu)
R
o
s
= shellside resistance adjusted to design (hr- ft
2
-F/Btu)

The test pressure is then adjusted to design reference conditions using the dimensionless number
of heat transfer units (NTU
o
) based on equation 5.2.16:

6-44
NTU
o
= (U
o
* A*) / (c*
p
* w*)
Where: NTU
o
= Number of Heat Transfer Units

o
= test value adjusted to design reference conditions
U
o
= test heat transfer coefficient corrected to design (Btu/hr- ft
2
- F)
A
o
2
)
c*
p
w* = flow rate (lb/hr)
The adjusted test pressure is then determined from saturation pressure associated with the
temperature calculated from equation 5.2.17:
T
o
s
= [(T
*
2
- T
*
1
) * e
- NTUo
] / (1 - e
- NTUo
)
Where: NTU
o
= Number of Heat Transfer Units

o
= test value adjusted to design reference conditions
T
o
s
= condenser steam saturation temperature (F)
T
*
2
T
*
1
The test adjusted steam pressure is then compared to the design pressure. If the adjusted test
pressure is below the design pressure, the condenser passes the acceptance test.
Discussion of Test Results
The objective of our testing program was to apply the methodology outlined in the new test code
especially those related to determining tube bundle fouling. The test provided reasonable results
and could be utilized as a demonstration of the test code methods. There was no acceptance
criteria placed on the retubing project however, the data indicated that the retubed condenser was
performing as expected.

6-45
As discussed above, the code recommended eight pairs of tubes be monitored to determine the
bundle fouling. Because this is a two-pass condenser we located two sets on each of the two
first pass outlet tubesheets and two sets on each of the second pass tubesheets. Since the
condenser has two passes, the second pass inlet was measured as well. This configuration
provided the determination of four fouling resistances in the first pass and four in the second
pass. Following the retube, the eight cleanreference tubes were plugged to allow the tube
bundle and representative fouled tubes to build up a fouling resistance. Once the clean tubes
were unplugged during an outage, the fouling resistance could be determined from the
temperature difference between the fouled and clean tubes as explained previously.
In this case, it took nearly eighteen months for the opportunity of an outage to remove the plugs
due to the excellent availability of the unit. Fortunately the cooling water at the station is quite
clean and so the tube fouling rate is low. During the same outage, the condenser tubes were
cleaned with brushes.
The tube fouling monitoring system was inspected during that outage and found it to be in
excellent condition. The wiring was essentially located as it had been installed the prior year and
all temperature sensors were working. The data recorded for the calculations presented in this
paper was taken approximately 4 weeks after the condenser cleaning. For many condensers with
aggressive fouling this would be too long to wait for the testing, because the cleantubes would
likely reach a level of fouling similar to the fouledtubes used for the reference. At the time of
the data collection however, three of the eight temperature readings were faulty inside the
waterbox. This may have occurred during the condenser cleaning or from the impact of the
circulation water during startup. Based on a review of the standard deviation of the fouling
resistances, one set of readings indicated a strong possibility that one of the first pass outlet
temperatures should have been labeled as a second pass inlet temperature.. If we assume this to
have been the case, then we have five fouling resistance values after eighteen months of service.
The values of the fouling resistance (R
+
f
) based on equation 5.1.11 (shown on in the Test
Computation section) are:

Location R
+
f
- foul resistance
1
st
Pass East Side (Upper) .0004928
1
st
Pass West Side (Upper) .0005217
1
st
Pass West Side (Upper) .0005405
2
nd
Pass East Side (Lower) .0005765
2
nd
Pass West Side (Lower) .0006823
Average .0005628
Each fouling resistance value compares fairly well to the average resistance as indicated by the
relatively low standard deviation. Therefore we had a higher level of confidence that the average
fouling resistance is representative of the tube bundle fouling.

6-46
A comparison of the fouling resistance to the HEI derived percent clean was made to get another
perspective on the measured test performance and its credibility. As follows, the estimate
indicated a tube bundle cleanliness of 72.5%. That was judged to be a reasonable value for this
condenser in service about 4 weeks after the tube cleaning and in service for 18 months.
R
fouled
in terms of HEI % Cleanliness:
R
fouled
= 1/U
HEI
+ R
+
f average

U
HEI
= 675 Btu/hr- ft
2
- F (100% clean at test conditions)
R
+
f average
= .0005628 hr- ft
2
-F/Btu
R
fouled
= 1/ U
HEI
+ R
+
f average
=1/675 + .0005628
R
fouled
= 0.0020442 hr- ft
2
-F/Btu
U
fouled
= 1/ R
fouled
= 1/0.0020442 = 489 Btu/hr- ft
2
- F
%c = U
fouled
/ U
HEI
= 489/675 = 72.5%
The test indicated that the overall test pressure adjusted to design was 2.39 in. HgA. The test
condenser pressure adjusted to design conditions was determined from the steam saturation
temperature calculated by applying the number of heat transfer units at the design conditions
(from equation 5.2.17 in the Code). We had predicted that the new design pressure for the
change in tube material and gauge to be 2.49 in. HgA based on HEI methods. Based on the
results of the performance test, the adjusted design pressure is 0.10 in. HgA better than our
expected design value and thus is a reasonable result compared to our expected value.
Methods of Fouling Monitoring
Most of the condenser performance degradation that occurs during operation can be blamed on
tube fouling. To maintain maximum station generation for the specific load and inlet water
temperatures encountered throughout the year and to decide when it is cost effective to take any
tube cleaning actions, it is very important to know the level of tube fouling and its impact on that
generation. With these considerations in mind and in addition to the performance monitoring
methods presented in the ASME PTC 12.2 1998 test code Appendix C, two methods of
monitoring fouling based on utilizing the tube fouling monitors are also offered.
The first method involves plugging several sets of tubes, cleaned to an as new condition, to act as
cleanreference tubes. The number of clean tubes required would depend on the period of time
that the monitoring program is to cover. Additionally, the number of tube fouling monitors
required may have to be based on experience with the program since fouling conditions vary
greatly depending on the quality of the circulating water.

6-47
The basis of the program is to monitor fouling using an initial set of clean reference tubes while
the other clean tubes remain plugged. The remaining plugged tubes will later be unplugged in
sets to act as new reference tubes. Depending on the aggressiveness of the fouling, the program
would likely require access to the condenser at least monthly to unplug clean tubes.
Initially the condenser would be cleaned and the reference tubes would be replaced with new or
cleaned to an as new condition. Several clean tubes would be plugged and the condenser
allowed to operate for a short period of time before removing the first set of plugged clean tubes.
The period of time will depend on the aggressiveness of the fouling, for example, a week or so
for mild organic or crystalline based fouling.
A baseline test to determine the fouling resistance is conducted. The condenser is then
monitored on a periodic basis to determine the fouling resistance. The fouling resistance is
compared to the baseline to determine the rate of fouling. The above method of converting the
fouling resistance to the HEI % cleanliness may be used if desired.
At some point, the fouling of the clean tubes will become similar to the fouled tubes. This can
be determined when the original temperature rise determined from the baseline test decreases
and the fouling resistance appears to decrease. When this condition is reached, the next set of
clean tubes should be unplugged to provide a new comparison to the reference.
Another method of monitoring fouling is to determine a baseline set of performance data and
treat the computed shellside resistance, metal resistance and tubeside resistance values as
constants for future fouling tests. As described previously and in the test code, calculate a
baseline overall test heat transfer coefficient (U
+
), fouling resistance (R
+
f
), metal resistance (R
+
m
),
tubeside resistance (R
+
t
), test shell side resistance (R
+
s
).
During future tests to monitor fouling the current overall test heat transfer coefficient (U
+
) is
determined. The metal resistance (R
+
m
), tubeside resistance (R
+
t
), test shell side resistance (R
+
s
)
can be treated as constants because they would not be expected to change with fouling. The
fouling resistance (R
+
f
) is determined by solving equation (5.1.12) for R
+
f
, as follows:
R
+
f
= 1 /U
+
-

R
+
m
- R
+
t
(d
o
/d
i
) - R
+
s
The fouling resistance can thus be monitored over time without depending on the reliability of
the fouling monitors.
Another benefit of the data provided from the tube fouling monitors is that depending on the
condenser discharge temperature instrumentation and temperature stratification the tube fouling
monitors may be a better indication of the average outlet condenser temperature.
Conclusions and Recommendations
The results of our test demonstration were consistent with the expectations determined from the
reasonable test results. As discussed in the Test Results section, the fouling resistance values of
this test exhibited a small standard deviation and when translated into an apparent cleanliness,
indicated a credible fouling test result or an acceptable level of fouling.

6-48
The condenser was retubed with a change in material and tube gauge to 20 BWG, 90-10 Cu-Ni.
HEI and the condenser manufacturers manual were used to predict the new design condenser
pressure of 2.49 in. HgA. In this case, the test measurements adjusted to design conditions
indicated a test pressure of 2.39 in. HgA and thus constituted a reasonable performance level
after the retubing. Of course, had this been an actual acceptance test for either a modular
rebundling or a new condenser, the test procedures would have been more stringent and included
a circulating water flow measurement.
The test code lists procedures to monitor condenser performance. We have offered additional
methods to monitor tube bundle fouling based on the calculation procedures and methods
presented in the test code.
Some precautions regarding a tube bundle fouling monitoring program should be offered. For
the program to be successful it is important to have access to the condenser should maintenance
of the fouling monitors be required.
Finally, we believe that our experience with the new ASME condenser test code PTC 12.2
1998, indicates that it provides practical and reasonable cost methods of determining the effect of
fouling and the overall performance of a condenser. The methods appear to be satisfactory for
performing acceptance testing and condenser performance monitoring.
References
1. ASME PTC 12.2 1998, Performance Test Code on Steam Surface Condensers, 1998.
2. Heat Exchange Institute Standards for Steam Surface Condensers, Ninth ed., 1995.
2002 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.
EPRI. ELECTRIFY THE WORLD is a service mark of the Electric
Power Research Institute, Inc.
Printed on recycled paper in the United States of America
1004116
Target:
Coal Boiler Performance Optimization and
Combustion NO
x
Control
EPRI 3412 Hillview Avenue, Palo Alto, California 94304 PO Box 10412, Palo Alto, California 94303 USA
800.313.3774 650.855.2121 askepri@epri.com www.epri.com
About EPRI
EPRI creates science and technology solutions for
the global energy and energy services industry. U.S.
electric utilities established the Electric Power
Research Institute in 1973 as a nonprofit research
consortium for the benefit of utility members, their
customers, and society. Now known simply as EPRI,
the company provides a wide range of innovative
products and services to more than 1000 energy-
related organizations in 40 countries. EPRIs
multidisciplinary team of scientists and engineers
draws on a worldwide network of technical and
business expertise to help solve todays toughest
energy and environmental problems.
EPRI. Electrify the World

Condenser Technology

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Condenser Technology

Uploaded by

Copyright:

Available Formats

Condenser Technology

Seminar and Conference

Stainless, 27-4-2, by K. E. Pinnow

N08367 45 45C (113 F) 18000

S44660 46 52 C (125 F) 20000+

, on-line mechanical cleaning was synonymous with the

Rocket tube cleaners to restore condenser performance in lieu of an off-line

You might also like