Comprehensive nuclear materials 5 02 water chemistry control in LWRs

5.02

Water Chemistry Control in LWRs

C. J. Wood
Electric Power Research Institute, Palo Alto, CA, USA

ß 2012 Elsevier Ltd. All rights reserved.

5.02.1

Introduction

18

5.02.2
5.02.2.1
5.02.2.2
5.02.2.3
5.02.2.4
5.02.2.5
5.02.3
5.02.3.1
5.02.3.2
5.02.3.3
5.02.3.4
5.02.4
5.02.4.1
5.02.4.2
5.02.4.3
5.02.4.4

5.02.5
5.02.6
References

BWR Chemistry Control
Evolution of BWR Chemistry Strategies
Mitigating Effects of Water Chemistry on Degradation of Reactor Materials
Radiation Field Control
Fuel Performance Issues
Online Addition of Noble Metals
PWR Primary Water Chemistry Control
Evolution of PWR Primary Chemistry Strategies
Materials Degradation
PWR Radiation Field Control
Fuel Performance
PWR Secondary System Water Chemistry Experience
Evolution of PWR Secondary Chemistry Strategies
Chemistry Effects on Materials Degradation of SGs
Control of Sludge Fouling of SGs
Lead Chemistry
Chemistry Control for FAC in BWRs and PWRs
Water Chemistry Control Strategies

19
19
20
23
26
27
27

27
29
33
35
37
37
40
43
44
45
45
46

Abbreviations
AO

AVT

BOP
BRAC

BWR
CGR
CRUD
DMA
DZO
EBA
ECP

Axial offset, referring to localized

flux depression in reactor core
caused by buildup of boroncontaining deposits. Originally
called AOA for axial offset anomaly.
All-volatile treatment, suing
ammonia for pH control in steam
generators
Balance of plant
BWR radiation and control,
referring to designated standard
points in BWR reactors for
radiation field measurements
Boiling water reactor
Crack growth rate
Corrosion product deposits on
fuel element surfaces
Dimethylamine
Depleted zinc oxide (BWRs)
Enriched boric acid (PWRs)
Electrochemical corrosion
potential

ETA
FAC
GE

HWC
HWC-L
HWC-M
IGA
IGSCC

LWR
MOX
MPA
MRC
NDE
NMCA
NWC
OD IGA/SCC

Ethanolamine
Flow-assisted corrosion
General electric, the vendor for
BWRs in the United States and
some other countries
Hydrogen water chemistry
HWC (low) with 0.2–0.5 ppm
hydrogen
HWC (moderate) with 1.6–2.0 ppm
hydrogen
Intergranular attack
Intergranular stress corrosion
cracking
Light water reactor
Mixed oxide fuel
3-methoxypropylamine
Molar ratio control (PWR
secondary side)
Nondestructive examination
Noble metal chemical addition
Normal water chemistry (BWRs)

Outside diameter IGA/SCC in
steam generator tubes

17


18

Water Chemistry Control in LWRs

OLNC
OTSG
PAA
PbSCC
PWR
PWSCC
SCC
SG
SHE

On-line noble chemistry
Once through steam generator
Poly acrylic acid
Lead assisted stress corrosion
cracking
Pressurized water reactor
Primary water stress corrosion
cracking
Stress corrosion cracking
Steam generator

Standard hydrogen electrode (for
ECP measurements)

5.02.1 Introduction
Other chapters of this comprehensive describe the various degradation processes affecting the structural materials used in the construction of nuclear power plants
(see Chapter 5.04, Corrosion and Stress Corrosion
Cracking of Ni-Base Alloys; Chapter 5.05, Corrosion and Stress Corrosion Cracking of Austenitic
Stainless Steels; and Chapter 5.06, Corrosion and
Environmentally-Assisted Cracking of Carbon and
Low-Alloy Steels). This chapter describes the influence
of water chemistry on corrosion of the most important
materials in light water reactors (LWRs). In particular,
alloys susceptible to intergranular attack (IGA) and stress
corrosion cracking (SCC) are significantly impacted
by water chemistry, most notably, sensitized 304 stainless steel in boiling water reactors (BWRs) and nickelbased alloys in pressurized water reactors (PWRs).
Excellent water quality is essential if material
degradation is to be controlled. In the early days of
nuclear power plant operation, impurities in the
coolant water were a major factor in causing excessive
corrosion. Chlorides and sulfates are particularly
aggressive in increasing intergranular stress corrosion
cracking (IGSCC) and other corrosion processes.
Transient increases of impurities in the coolant that
occur during fault conditions (e.g., condenser leaks
and ingress of oil or ion exchange resins) proved to be
particularly damaging. Thus, water chemistry was
traditionally regarded as a key cause of material degradation. Initial efforts to improve water quality
brought about a slow but steady reduction in impurities through improved design and operation of purification systems. Not only were the average
concentrations of impurities reduced over time, but
the frequency and magnitude of impurity ‘spikes’

from transient fault conditions were also diminished.

However, excellent water chemistry alone was not
sufficient to control corrosion. Hence, programs to
modify water chemistry were introduced, including
minimizing oxygen to reduce the electrochemical
corrosion potential (ECP) in BWRs, and oxygen and
pH control in PWRs. More recently, additives to
further inhibit the corrosion process have been developed and are now in widespread use. As a result,
water chemistry advances are now an important part
of the overall operating strategy to control material
degradation.
Primary system water chemistry also affects fuel
performance through the deposition of corrosion
products on fuel pin surfaces, and influences radiation fields outside the core. Core uprating through
increased fuel duty has reduced margins for tolerating corrosion products (CRUD) on BWR fuel pin
surfaces. In PWRs, increasing fuel cycle duration
has increased the challenge of controlling pH within
the optimum range. At the same time, regulatory
limits on worker radiation exposure are tending to
be tightened worldwide, putting pressure on the
operators to reduce radiation dose rates. Successful
operation of PWR steam generators (SGs) and the
remainder of the secondary system demand strict
water chemistry control in secondary side systems if
corrosion problems are to be avoided.
Other operating parameters also influence the
optimization process, for example, life extension (to
60 years) has emphasized the importance of controlling
degradation of circuit materials. Therefore, although

control of structural material degradation remains the
highest priority, water chemistry must be optimized
between the sometimes-conflicting requirements affecting other parts of the reactor.
Advances in water chemistry have enabled plant
operators to respond successfully to these technical
challenges, and the overall performance has steadily
improved in recent years.1 Plant-specific considerations sometimes influence or indeed limit the options
for controlling water chemistry, so we see different
chemistry specifications at different plants. This is especially true internationally and significant differences
between countries are noted. The US industry started
developing water chemistry guidelines 25–30 years
ago, and these now provide the technical basis for
guidelines in many other countries. The early editions
of these guidelines presented impurity specifications
and required action if limits were exceeded. When
advanced water chemistries were developed and qualified, the guidelines evolved into a menu of options
within an envelope of specifications that should not be


Water Chemistry Control in LWRs

exceeded. Guidance is now provided on how to select
a plant-specific water chemistry strategy.2
The basis for water chemistry control was discussed in detail by Cohen.3 The remainder of this
chapter describes more recent water chemistry developments for BWRs, PWR primary systems, and
PWR secondary systems including SGs, with a
short section on flow-assisted corrosion (FAC) in
both BWRs and PWRs.

5.02.2 BWR Chemistry Control

5.02.2.1 Evolution of BWR Chemistry
Strategies
BWR water chemistry has to be optimized between
the requirements to minimize material degradation,
avoid fuel performance issues, and control radiation
fields. These factors are depicted in Figure 1,4 which
also includes the main chemistry changes involved in
the optimization process.
Plant-specific considerations sometimes influence
or indeed limit the options for controlling water chemistry, so we see different chemistry specifications at
different plants. This is especially true internationally
and significant differences in chemistry strategies
between countries are noted. Design features are an
important reason for these different chemistry regimes,
to which must be added the effects of different operational strategies in recent years. For example, a key
issue facing BWRs in the United States concerns
IGSCC of reactor internals, as discussed in other
chapters. The occurrence of IGSCC resulted in the

Clad corrosion
crud deposition:
Limits on
feedwater zinc

Impurity control:
Monitoring/analysis
required

implementation of hydrogen water chemistry, with or
without noble metal chemical addition (NMCA), to

ensure that extended plant lifetimes are achieved.
German plants use 347 stainless steel, which is less
susceptible to IGSCC than sensitized 304 stainless
steel used originally in US-designed plants. Some
Swedish and Japanese plants have replaced 304 stainless steel reactor internals with 316 nuclear grade
material to minimize potential problems, as this material is less susceptible to IGSCC. As a result, many of
these plants continue to use oxygenated normal water
chemistry, whereas all US plants control IGSCC
through the use of hydrogen water chemistry (HWC)
with or without normal metal chemical addition to
improve the efficiency of the hydrogen in reducing
ECP. Second, BWRs in United States undoubtedly
have greater cobalt sources than plants in most other
countries, despite strong efforts to replace cobalt
sources. This resulted in higher out-of-core radiation
fields, leading all US plants to implement zinc injection
to control fields, whereas only a small number of plants
of other designs use zinc. Third, the move to longer fuel
cycles and increased fuel duty at US plants, while
having major economic benefits, has led to new constraints on chemistry specifications in order to avoid
fuel performance issues.
Figure 2 depicts the changing chemistry strategies over the past 30 years, showing the focus on
improving water quality in the early 1980s and the
move to educing chemistry to control IGSCC, which
in turn resulted in increased radiation fields, subsequently controlled by zinc injection.

Materials
degradation
and mitigation

Water
chemistry
guidelines

Fuel
performance

Chemistry
control issues

Figure 1 Boiling water reactor chemistry interactions.

19

BWR internals
IGSCC, IASCC:
HWC or NMC
required

Radiation
exposure

Radiation fields
crud bursts:
Zinc required


20

Water Chemistry Control in LWRs

Increasing concerns about core internals cracking led to the need to increase hydrogen injection
rates, which in turn resulted in the introduction
of NMCA to reduce operating radiation fields from
N-16. Figure 3 shows the rate of implementation of
HWC, zinc and NMCA, and online noble metal
addition (OLNC). The rationale and implications of
these developments are discussed in greater detail in
subsequent sections.
The goal for BWRs is therefore to specify chemistry
regimes that, together with the improved materials
used in replacement components (e.g., 316 nuclear
grade stainless steel), will ensure that the full extended
life of the plants will be achieved without the need for
further major replacements. At the same time, radiation dose rates, and hence worker radiation exposure,
must be closely controlled, and fuel performance must
not be adversely affected by chemistry changes.

The first requirement of plant chemistry is to maintain high-purity water in all coolant systems, including
the need to avoid impurity transients, which are
beyond the scope of this paper. The performance of
all plants has improved steadily over the years, as
shown by the trend for reactor water conductivity for
GE-designed plants, given in Figure 4. This figure
shows that conductivity now approaches the theoretical minimum for pure water. In fact, deliberately added
chemicals, such as zinc (discussed in the following
section), account for much of the difference between
measured values and the theoretical minimum.
The conductivity data are consistent with the
reactor water concentrations for sulfate and chloride.

In fact, sulfate is the most aggressive impurity from
the viewpoint of IGSCC, and much effort has gone
into reducing it.
5.02.2.2 Mitigating Effects of Water
Chemistry on Degradation of Reactor
Materials

1977: Neutral,
oxygenated water

Corrosion, radiation buildup issues

1980s: Purer is better

IGSCC was first observed in small bore piping using
sensitized 304 stainless steel fairly soon after BWRs
started operation. Laboratory studies showed that
impurities increased IGSCC rates, and in fact water
quality in BWRs gradually improved in the early
1980s. However, the same studies found IGSCC in
high-purity oxygenated water typical of good BWR
operations. The key parameter affecting IGSCC was
found to be ECP, as shown in Figure 5. In this laboratory test, the change from oxidizing conditions
typical of normal water chemistry (NWC) operation

Chemistry guidelines

Late 1980s–1990s:
HWC, zinc

Controlling IGSCC,
radiation buildup

2000s: Noble metal
chemical addition

Core internals cracking
control with lower fields

Promising new option

2006–2008:
Online Noblechem

Figure 2 Evolution of Boiling water reactor chemistry
options from 1977 to 2008.

40

Number of BWRs

35

Zn injection

NMCA

HWC (no NMCA)

OLNC

30
25
20
15
10
5
0
1983

1988

1993

1998

2003

2008

Figure 3 Implementation of zinc injection, hydrogen water chemistry, noble metals chemical addition, and online noble
metal at US boiling water reactors.


Water Chemistry Control in LWRs

21

0.40
0.35

EPRI action level 1

Conductivity ( µS cm–1)

0.30
0.25
0.20
0.15
0.10
0.05
Theoretical conductivity limit, 25 ºC

0.00
1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

Figure 4 Boiling water reactor mean reactor water conductivity at US boiling water reactor.

250

0.4950

0.4945
200

2.7 ϫ 10−8 mm s–1
1 ϫ 10−6
mm s–1

150

0.4935

0.4930

0.4925

100
CT2 #7-304SS 4 dpa
Constant load, 19 ksi√in.

Dissolved O2

Outlet cond: 0.30 μS cm–1

50

Inlet cond: 0.27 μS cm–1 Na2SO4

0.4920

0.4915
1488

Dissolved oxygen (ppb)

Crack length (in.)

0.4940

1508

1528

1548

1568

1588

0
1608

Test time (h)
Figure 5 Laboratory results showing the effect of reducing oxygen concentration on crack growth of 304 stainless steel.

to reducing conditions greatly reduced the rate of
crack growth.
Furthermore, hydrogen injection was effective at
reducing the ECP in BWRs, as shown in Figure 6.
In this figure, it can be seen that crack growth rates
(CGR) for Alloy 182 were low in hydrogen water

chemistry (HWC), but increased greatly when the
plant reverted to normal water chemistry (NWC).
These results indicated that continuous hydrogen
injection was required to fully mitigate cracking.
Examination of extensive inspection data from several
plants indicated that no IGSCC was observed with an


22

Water Chemistry Control in LWRs

901.00
900.00

HWC
ECP = −510 mV (SHE)

NWC
ECP = +110 mV (SHE)

Crack length

174 miles year-1

HWC

< 5 miles year-1

899.00
898.00
897.00
< 5 miles year -1

Alloy 182

896.00
895.00
800

900

1000

1100

1200

1300
Time (h)

1400

1500

1600

1700

Figure 6 Effect of hydrogen water chemistry on crack growth of Alloy 182.

ECP of À230 mV or lower, using a standard hydrogen
electrode (SHE). This is the basis for the À230 mV
requirement used by US plants for IGSCC control.
In BWRs, the radiation field in the core decomposes
water to hydrogen and oxygen species, most of which
immediately recombine back to water. But some
remain as oxygen or hydrogen peroxide, because some
hydrogen is stripped into the steam phase before it can
recombine. These same radiolysis reactions cause

hydrogen to react with oxygen or peroxide to reduce
ECP. These reactions occur mainly in the downcomer,
and relatively low hydrogen concentrations are effective at lowering ECP in out-of-core regions of the
system. More than half the BWRs in the United States
adopted low hydrogen injection rates of 0.2–0.5 ppm
(called HWC-L), which, coupled with the replacement
of recirculation piping using 316 stainless steel, mitigated IGSCC of recirculation piping.
In the 1990s, concerns about the cracking of core
internals increased, but the low concentrations of
hydrogen used to protect out-of-core regions were
not sufficient to reduce ECP enough to mitigate
IGSCC of in-core materials, because of the radiolysis
of water occurring in the core. As a result, it was
necessary to increase hydrogen concentrations to
1.6–2.0 ppm to lower the in-core ECP sufficiently to
provide protection in the reactor vessel (termed
HWC-M for moderate concentrations of hydrogen).
Although this approach was effective in protecting
core internals, it also increased radiation fields in the
steam side of the circuit, including the turbines, as a
result of carryover of nitrogen-16 under reducing
chemistry. (Under the oxidizing conditions of NWC,
most of the N-16 remains in the water as soluble

species such as nitrate, and only a small percent is
transported with the steam.) In some plants, local
shielding of turbine components has reduced the
impact of the gamma radiation to acceptable levels,
but the projected 4–6-fold increase did in fact curtail
plans for increased hydrogen injection rates at many

plants. Note that these N-16 radiation fields are a
problem only when the plant is at power, as rapid
decay occurs at shutdown because of the short halflife of N-16. (By contrast, out-of-core radiation fields
from Cobalt-60 persist after shutdown and impact on
maintenance work during outages.)
NMCA was developed to increase the efficiency
of hydrogen in BWR cores, to avoid high N-16 fields.
In this process, a nanolayer of platinum þ rhodium is
deposited on the wetted surfaces of the reactor. These
treated surfaces catalyze the hydrogen redox reaction, converting oxygen back to water. When the
addition of hydrogen to the feedwater raises the
molar ratio of H2 to O2 to 2 or higher, the ECP of
the treated surfaces drops to the hydrogen/oxygen
redox potential, which is about À450 mV (SHE). This
can be achieved with hydrogen concentrations of
only about 0.2 ppm, and under these conditions, the
main steam radiation level is not increased to an
unacceptable level. The first plant used NMCA successfully in 1997, and over 25 plants have already
followed, with excellent results. Field measurements
show that NMCA has been effective in providing
mitigation against IGSCC by lowering the ECP
below the À230 mV (SHE) threshold with relatively
low hydrogen injection rates.
The NMCA process is typically applied at refueling outage, before new fuel is inserted into the core,


Water Chemistry Control in LWRs

additional benefit with NMCA on the upper, outer
shroud regions, as indicated by the additional shading

in the left-hand side of the figure5. It is estimated that
noble metals protect slightly more of the outer core
region than does moderate HWC (HWC-M), but the
difference is not significant.
Figure 8 shows the dramatic benefit of noble
metals in reducing the rate of stub tube cracking at
Nine Mile Point 1 since the application in 2000.
Before 2000, several stub tubes had to be repaired
or replaced at each outage, but since the application,
only one tube leaked, and this was believed to have
already cracked before NMCA.
Recently, attention has been focused on the online
application of noble metals, with the first application
at the KKM plant in Switzerland. By April 2008,
there were four applications in the United States.
This is discussed in a later section.

HWC protected regions

NMCA protected regions

and is effective for about three fuel cycles, before
reapplication is necessary. The regions of the reactor
vessel internals that are protected by HWC-M or
NMCA are shown in Figure 7. While both techniques offer significant areas of mitigation, there is an

5.02.2.3

Radiation Field Control

Corrosion products deposited on the fuel become
activated, are released back into the coolant, and
may be deposited on out-of-core surfaces. Both soluble and insoluble species may be involved, the latter
tending to deposit in stagnate areas (‘crud traps’). The
chemistry changes to control IGSCC resulted in
increased out-of-core radiation fields, and the implementation by most plants of depleted zinc injection to

Figure 7 Mitigated regions of the boiling water reactor
core.

Number of stub tubes identified with IGSCC
throughwall cracking based on leakage

12

10

8

Noble metal applied
mid cycle
may 2000

6

4

2

0

1984–1985 1986–1987 1988–1990

23

1991

1993

1995

1997

1999

2001

2003

2005

2007

RFO-11

RFO-12

RFO-13

RFO-14

RFO-15

RFO-16

RFO-17

RFO-18

RFO-19

Year
Figure 8 Mitigation of stub tube cracking at Nine Mile Point Unit 1.


24

Water Chemistry Control in LWRs

control dose rates, as discussed later in this section.
During shutdowns, the major radiation source for
personnel exposure is activated corrosion products,
deposited on primary system surfaces. Exposures are
generally accumulated at high-radiation field locations
where maintenance work is frequently needed.
Although improvement of maintenance equipment
and procedures, reduction of maintenance requirements, increased hot-spot shielding, and control of
contamination dispersion have significantly reduced
total exposure, further reduction of radiation fields is
a major goal in programs for minimizing occupational
radiation exposure.

The primary source of radiation field buildup on
out-of-core surfaces in BWRs is 60Co, which in
mature plants usually accounts for 80–90% of the
total dose. 60Co has a relatively long half-life of
5.27 years. The higher the soluble 60Co concentration
in the coolant, the more 60Co is incorporated and
deposited on out-of-core systems and components,
resulting in higher dose rates on recirculation piping,
the reactor water cleanup system, dead legs, and
other crud traps in the system. Other activated transition metals such as 54Mn, 58Co, 59Fe, and 65Zn
contribute the remainder of the dose. 51Cr also contributes significantly to the piping dose in some
NMCA plants. The radiation fields commonly
measured in a BWR at the straight vertical section
of recirculation pipes are considered to be more
representative for the purposes of radiation buildup
trending and comparison with other plants. These
measurements are done in a prescribed manner
developed under the EPRI BWR Radiation and Control program and are called BRAC point measurements. These measurements represent primarily the
incorporation of soluble 60Co into the corrosion film
on the piping surfaces and tend to be a fairly good
predictor of drywell dose rates. The deposition of
particulate oxides that contain 60Co and other activated species can also contribute significantly to outof-core radiation levels in BWRs, especially in hot
spots. The particulate oxides, which vary in size,
originate primarily from corrosion of the steam/condensate system and are introduced via the feedwater.
The sole precursor of the gamma-emitting 60Co
isotope is 59Co. 59Co is present as an impurity in the
nickel in structural alloys used in BWRs (e.g., Type
304 stainless steel) and is the main constituent of
wear-resistant alloys (e.g., Stellite), used as hard facing in valves and other applications requiring outstanding wear resistance. Corrosion and wear lead to
release of 59Co into the coolant from these sources,

which is transported to the core and incorporated
into the crud that deposits on the fuel rods. The
59
Co is activated to 60Co by neutron activation,
released back into the coolant, and incorporated as
a minor constituent into the passive films that form
on components that are inspected, repaired, and
replaced by maintenance personnel. Components in
the neutron flux (e.g., the control blades) directly
release 60Co. Cobalt source removal is clearly important if radiation fields are to be minimized. Another
gamma-emitting isotope, 58Co, is formed by the activation of nickel from stainless steel and nickel-based
alloys. 58Co has a shorter half-life and is not as major
a contributor to radiation fields as 60Co in BWRs, but
is much more significant in PWRs.
Shutdown drywell dose rates increase when coolant chemistry is changed for the first time from
oxidizing (NWC) to reducing (HWC) conditions.
This results from a partial restructuring of the oxides
formed under the oxidizing conditions of NWC
(Fe2O3 type) to a more reducing spinel type oxide
compound (Fe3O4 type). The oxides affected are the
fuel deposits, the corrosion films on stainless steel
piping, and out-of core deposits. This results in an
increase in the chemical cobalt (and 60Co) concentration in the oxide because of the higher solid-state
solubility of transition metals in the spinel structure.
The presence of a higher soluble reactor 60Co concentration released from fuel crud while this conversion is occurring only aggravates the situation. The
processes are depicted in Figure 9. The net result at
most plants is a temporary increase in reactor water
60
Co, both soluble and insoluble forms, which leads to

significantly increased shutdown dose rates because
of both the increased reactor water concentrations
and the increased capacity for transition metal uptake
by the spinel phases.6

Oxide stable
under normal
water chemistry
Fe2O3
(containing 60Co,
58Co, 54Mn, etc.)

• Corrosion films
• Vessel crud
• Fuel crud

Restructuring under
HWC conditions

Fe3O4
form of oxide

Small insoluble
particles containing
60Co, 54Mn, etc.
Soluble 60Co, etc.
released during
restructure

Figure 9 Boiling water reactor oxide behavior under

reducing conditions.


Water Chemistry Control in LWRs

25

0.8
Before Zn addition
After Zn addition

RxW 60Co (Ci kg −1)

0.6

0.4

0.2

0
Brunswick-1

Brunswick-2

Dresden-2

Figure 10 Hydrogen water chemistry plant RxW

60

Duane Arnold

Fitz patrick

Monticello

Pilgrim

Co response to zinc addition.

As mentioned earlier, zinc addition reduces radiation field buildup. The mechanism of the zinc ion effect
is complex, as release of 60Co from fuel crud is reduced,
and deposition out-core is also reduced. Overall, reactor water 60Co is decreased significantly after zinc
addition, as shown by plant data in Figure 10.
Aqueous zinc ion promotes the formation of a
more protective spinel-structured corrosion film on
stainless steel, especially when reducing conditions
are present. Second, both cobalt and zinc favor tetrahedral sites in the spinel structure, but the site preference energy favors zinc incorporation. Thus, the
available sites have a higher probability of being filled
with a zinc ion than a cobalt ion (or 60Co ion), and
hence the uptake of 60Co into the film will be significantly less if zinc ion is present in the water. The
60
Co remains longer in the water and is eventually
removed by the cleanup system.
The zinc was originally added to the feedwater as
ZnO, but it was quickly found that the 65Zn that was
created by activation of the naturally occurring 64Zn
isotope in natural zinc created problems. With the
use of zinc oxide depleted in the 64Zn isotope, called
depleted zinc oxide (DZO), this drawback was eliminated. Because of the high cost of DZO, feedwater

zinc injection was not implemented widely until
HWC shutdown dose issues emerged.
For the case of plants treated with NMCA and
injecting hydrogen, the oxidant concentration on the
surface of the stainless steel is zero (due to the Pt

and Rh catalyzing the reaction of any oxidant with
the surplus hydrogen). The net result is that the ECP
is at or very near the hydrogen redox potential,
typically about –490 mV (SHE) for neutral BWR
water. This low potential causes a much more thorough restructuring of the oxides to the spinel state
than observed under moderate hydrogen water
chemistry (HWC-M).
Feedwater iron ingress has a significant influence
on the effectiveness of zinc injection. As discussed in
the next section, deposits on fuel cladding surfaces
(called ‘CRUD’) are mainly composed of iron oxides,
with other constituents. Therefore, reducing iron
ingress from the feedwater has the benefit of minimizing crud buildup, which is important for fuel
reliability (next section). For these reasons, extensive
efforts have been made to reduce iron ingress, with
significant success. Furthermore, fuel crud has a large
capacity for incorporating zinc and is in fact where
most of the zinc ends up. The lower the amount of
crud on the fuel, the greater the proportion of zinc
that remains in solution and can subsequently be
incorporated in out-of-core surfaces. Therefore, at
plants with low feedwater iron, less zinc is captured
by the crud on the fuel, so a relatively greater amount
remains in solution and is available to control out-of

core radiation fields. This is very important, as zinc
injection rates are limited by fuel performance concerns, and hence lowering feedwater iron is essential
for maintaining lower radiation fields.


26

Water Chemistry Control in LWRs

5.02.2.4

Fuel Performance Issues

Fuel durability has improved over the years, and
failures have declined, helped by improvements in
water purity. In operation, zircaloy fuel cladding
develops a thin oxide layer (ZrO2), which typically
does not adversely affect performance. However, an
increase of deposition of corrosion product deposits
(‘crud’) on this oxide film is undesirable because it
can reduce heat transfer and increase fuel pin temperatures, with resultant increased corrosion of the
fuel cladding, ultimately increasing the risk of fuel
failure. Moreover, the addition of additives to control
corrosion may increase the risk of crud buildup on
the fuel. For example, zinc and noble metals in BWRs
tend to increase the adherence of crud deposits on
the fuel, which can result in undesirable oxide spalling in higher-rated cores. In fact, corrosion-related
fuel failures occurred at four plants in the United
States between 1999 and 2003. Although the precise
root cause of fuel failures is often difficult to determine, it is clear that excessive crud buildup played a

role in these failures. With progressive uprating of
fuel duty in both PWRs (and BWRs), the margin to
tolerate crud has been reduced and additional care
has to be taken in specifying the water chemistry to
avoid undesirable fuel performance issues. Despite
these more demanding conditions, fuel failures have
decreased in recent years.
Concern about the possibility of adverse effects of
NMCA on fuel has prompted imposition of a strict
limit on the amount of noble metal that can end up on
the fuel and guidance on the injection of zinc. Plant
data indicate that spalling of the corrosion layer from

fuel cladding, which is often regarded as a precursor
to cladding failure, is prevented if the cycle average
feedwater zinc is maintained below 0.4 ppb in NMCA
plants (0.6 ppb for non-NMCA plants). More recent
data indicate that quarterly averages may be as high
as 0.5 ppb for NMCA plants, without occurrence of
spalling.5
These feedwater zinc data are the basis for limits
in the water chemistry guidelines. The 2008 chemistry guidelines7 retain the cycle average feedwater
zinc limit of 0.4 ppb (0.6 ppb for non-NMCA plants)
but enable a slight increase in the quarterly average
to 0.5 ppb, which may allow flexibility in controlling
radiation buildup in parts of the cycle.
The tighter control of water chemistry in recent
years has been successful in controlling crud formation on fuel cladding, and Figure 118 shows failures
from pellet–clad interaction causing SCC, fabrication
defects, debris, and crud/corrosion. Note that there

have been zero crud/cladding related fuel failures in
US BWRs since 2004 (although assessment of 2007
failures is not yet complete, crud/corrosion is not
believed to be a factor here).
Analysis of recent plant data confirms that control
of feedwater iron ingress has the positive benefit of
reducing the amount of crud on the fuel. Control of
copper, which generally originates from admiralty
brass alloys, is also beneficial; not only can copper
have detrimental effects on the fuel, but it also limits
the ability of hydrogen to reduce the ECP, and it also
leads to higher radiation fields. As a result, most US
plants have replaced condensers containing brass
tubing.

Number of failed assemblies

30
25
20

PCI-SCC
Unknown
Fabrication
Debris
Crud/corrosion

15
10
5

0
2000

2001

2002

2003
2004
EOC year

2005

2006

Figure 11 US boiling water reactor fuel failures by mechanism for each end-of-cycle (EOC) year.

2007


Water Chemistry Control in LWRs

5.02.2.5

Online Addition of Noble Metals

As discussed earlier, the classic NMCA process is
generally applied during refueling outages before
the new fuel is loaded into the core. Reapplication
after about three cycles of operation takes approximately 2 days, while the plant maintains 107–154  C

as it enters the refueling outage. To reduce this outage
time, GE-Hitachi developed OLNC, first demonstrated at KKM (a GE design of plant in Switzerland)
in 2005, with several more additions subsequently.
Preliminary results indicate that there have been no
unexpected chemistry effects during the first OLNC
applications, and shutdown radiation fields actually
decreased at KKM after OLNC.5 Subsequently, CGR
of susceptible welds decreased significantly, as shown
by the decrease in slopes in Figure 12 after OLNC
initiation for two welds that have been monitored for
several years.
The effects of OLNC on fuel have been extensively studied in fuel removed from KKM, and no
adverse effects have been observed. The jury is still
out on this concern, but the general assessment is that
OLNC will have no more impact than the classic
application, and may well prove to be of less concern.
More IGSCC and fuel measurements are planned,
but with no issues emerging to date, it appears that
OLNC applications about every 12 months would be
effective and economical, avoiding the critical path
time necessitated for the classic NMCA application
during refueling outages. Initial OLNC applications
have been carried out at plants that had previously
applied noble metals in the classic off-power manner.

However, the first OLNC application at a plant that
has not used noble meals previously occurred in late
2008, but no results are available.

5.02.3 PWR Primary Water

Chemistry Control
5.02.3.1 Evolution of PWR Primary
Chemistry Strategies
In the very early days of PWR operation, heavy crud
buildup on fuel cladding surfaces was caused by the
transport of corrosion products from the SGs into the
reactor core. As a result, activated corrosion products
caused high-radiation fields on out-of-core surfaces
(Figure 13), fuel performance was compromised, and
even coolant flow issues were observed in some
plants.
These problems were initially mitigated by imposing
a hydrogen overpressure on the primary system, to
reduce the ECP, and raising the primary chemistry
pH. Materials degradation in primary systems was
then not a major concern, with most of the emphasis
focused on secondary side corrosion issues in the
SGs. Commercial PWR power plants use a steadily
decreasing concentration of boric acid as a chemical
shim (for reactor control) throughout the fuel cycle,
which results in the use of lithium hydroxide to
control pH. Some 30 years ago, the concept of
‘coordinated boron and lithium’ was developed,
whereby the concentration of LiOH was gradually
reduced in line with the boric acid reduction to
maintain a constant pH.

300
NC
appl.

HWC

Indication length (mm)

250
Ind 9

Ind10

Not inspected in 00, 01,04

200
150
100
50

27

OLNC
37 g
OLNC
98 g
OLNC
198 g
OLNC
199 g

Indications 9,10
may be seeing

mitigation by OLNC

0
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Year
Figure 12 Ultrasonic inspection results after online noble metal chemical addition.


28

Water Chemistry Control in LWRs

Corrosion products
deposit out of core
Pressurizer
Steam generator
Corrosion
products
activated in
reactor core

Corrosion products
released from SG tubing
Coolant pump
Reactor
Primary loop

Figure 13 Transport and activation of corrosion products in pressurized water reactor primary systems.

Corrosion products released from the steam generator tubes are transported, dissolved, or deposited

by the coolant on the basis of solubility differences.
The solubility of nickel and iron depend on pH,
temperature, and redox potential, all of which vary
with location around the nonisothermal system.
Originally, a constant at-temperature pH of 6.9 was
recommended, based on the minimum temperature
coefficient of solubility of magnetite. In fact, it was
determined that heavy fuel crud buildup was avoided
if a constant pH of at least 6.9 was maintained. This
was possible with 12-month fuel cycles, but fuel cladding corrosion concerns limited the maximum LiOH
concentration to 2.2 ppm. Consequently, plants often
started the fuel cycle with pH below 6.9, which
resulted in deposition of corrosion products on the
fuel, activation of cobalt and nickel, and subsequent
transport to out-of-core surfaces, resulting in radiation fields remaining relatively high.
Even though detailed studies of fuel crud showed
that the prime constituent of the crud was nickel
ferrite (for which the optimum pH is 7.4), this coordinated chemistry had remained the standard for
many years, until higher pHs became the norm in
the 1990s. Although research and plant demonstrations showed that the 2.2 ppm limit was excessively
conservative, the move to higher Li concentrations
has been slow. However, detailed fuel examinations
from a recent plant demonstration (that will be discussed later) have indicated that Li can be raised to as
high as 6 ppm.

About 25 years ago, primary water stress corrosion cracking (PWSCC) of Alloy 600 SG tubes was
observed in a few plants, leading to studies on mitigating this effect. Following successful demonstration of zinc injection in BWRs, initial field tests at
PWRs showed that radiation fields were reduced,
and laboratory studies indicating that PWSCC was
reduced were eventually confirmed. As a result, zinc

injection is being implemented at an increasing rate,
although concerns about fuel performance at highduty plants have not been completely resolved. Most
recently, buildup of boron-containing crud in areas
of subcooled nucleate boiling leading to localized
flux depression has encouraged the use of higher Li
concentrations to minimize corrosion product transport. Concerns about the potential adverse effects of
zinc deposited in high-crud regions have resulted in
several highly rated plants applying in situ ultrasonic
fuel cleaning before implementing zinc injection.
Although zinc injection was developed for radiation field control, laboratory studies showed that it
also inhibited SCC under PWR conditions. The
identification of PWSCC in reactor vessel penetrations in the last 15–20 years has encouraged the use
of zinc injection, but has also focused attention on
the effects of dissolved hydrogen, for which the
recommended range has remained 25–50 ml kgÀ1
for 30 years. It now appears that raising hydrogen
will reduce PWSCC rates, while lowering it may
delay initiation of PWSCC. The interactions of
materials, radiation fields, and fuels in PWR primary


Water Chemistry Control in LWRs

PWSCC:
pH (Li, B) minimal effect
Zn beneficial
dissolved H2 effect

Plant
operations

29

Dissolved H2
control range

Materials
degradation

PWR
chemistry
control

Fuel
performance

Plant
dose
rates
Radiation fields:
pH (Li, B), Zn
beneficial

Crud deposition:
Zn concern for
highly rated cores

Figure 14 Pressurized water reactor primary chemistry optimization. Reproduced from Fruzzetti, K.; Perkins, D. PWR
chemistry: EPRI perspective on technical issues and industry research. In VGB NPC’08 Water Chemistry Conference, Berlin,
Sept 14–18, 2008.

Dissolved (H2) range changes

Elevated constant pH (7.3/7.4)
Ultrasonic fuel cleaning
Elevated constant pH (7.1/7.2)
Zinc injection
Modified elevated lithium program
EPRI water chemistry guidelines
Elevated lithium program
Constant pH 6.9
1975

1980

1985

1990

1995

2000

2006 2008

Figure 15 Pressurized water reactor primary chemistry
changes at US plants.

chemistry and optimization issues covered in the
water chemistry guidelines, which are discussed

later, are depicted in Figure 14.
The evolution of water chemistry control in PWR
primary systems in the United States over the last
30 years is shown in Figure 15.
The following sections address the three main
factors – pH control, zinc injection, and dissolved
hydrogen control – that have dominated PWR primary chemistry strategies in the past and continue to
do so today.9 Each of these factors is considered from
the viewpoint of materials degradation, radiation
field control, and fuel performance concerns.

5.02.3.2

Materials Degradation

Materials degradation has been covered in detail in
Chapter 5.04, Corrosion and Stress Corrosion
Cracking of Ni-Base Alloys and Chapter 5.05, Corrosion and Stress Corrosion Cracking of Austenitic
Stainless Steels, and here only the specific effects of
water chemistry variables on materials in PWR primary systems will be reviewed, particularly those that
may affect the chemistry of optimization process.
Recent papers by Andresen et al.10,11 provide
detailed results of a comprehensive study of the effects
of PWR primary water chemistry on PWSCC of
nickel-based alloys. Extensive studies have been carried out to determine the effect of lithium, boron, and
pH on PWSCC, and the generally held conclusion is
that any effects are minimal, especially compared to
material susceptibility, stress state and temperature,
and other operational issues. Crack initiation tests
using the most reliable types of reverse U-bend specimens indicate that pH has a relatively small effect on

crack initiation (generally less than a factor of 2).
Although the most rapid crack initiation occurred at
pH310  C 7.25, with slower rates at higher or lower pHs,
CGR tests generally confirm that pH has minimal
effect.
The effect of lithium is even smaller than the pH
effect, and the influence of boron is minor or nonexistent. Andresen et al. concluded that the effects of relevant variations in PWR primary water chemistry (B, Li,


30

Water Chemistry Control in LWRs

and pH) have little effect on the SCC growth rate in
Alloy 600, and thus provide little opportunity for mitigation of PWSCC. Plant data have found no adverse
effects from increasing lithium and pH in primary
systems. As a result, it is considered that adjusting
pH, lithium, or boron to minimize crack initiation
may be of minimal value. The 2007 edition of the
PWR Primary Water Chemistry Guidelines12 reviewed
the most recent data and concluded that pH strategy
changes based on PWSCC considerations are not warranted. This means that plants have the flexibility to
pursue B/Li/pHt chemistry adjustments to minimize
crud transport and radiation buildup without concern
for negative effects on PWSCC susceptibility of nickelbased alloys, although of course chloride and sulfate
impurities should continue to be minimized.
Following good experience in BWRs, zinc injection has been implemented in the primary systems of
PWRs, both to reduce primary side cracking of nickelbased alloys and to control dose rates. The qualification
work for BWRs showed that zinc inhibited SCC, but
the benefit was not sufficient to avoid the need for

hydrogen water chemistry to mitigate IGSCC. Thus,
the motivation for BWR zinc injection was exclusively
radiation field control.
The situation in PWRs is different, as laboratory
work13 showed that initiation of PWSCC was significantly delayed by zinc injection, and hence the motivation for the initial applications of zinc in most US
PWRs at the 10–30 ppb level was to control PWSCC
of SG tubing. Additionally, German-designed PWRs
and a few US plants used $5 ppb depleted zinc for
radiation control.

Figure 16 shows the rate of introduction of zinc
injection at PWRs worldwide.
Zinc injection produces thinner, more protective
oxides on stainless steel and Alloy 600, with zinc
displacing Co2þ, Ni2þ, and Fe2þ from normal spinels
to give ZnCr2O4, which is very stable. The benefits
of PWR zinc injection have been clearly demonstrated in reducing PWSCC degradation (especially
growth rate) of Alloy 600 SG tubes, and in controlling
radiation fields. Evaluation of currently available
laboratory data2 indicates that PWSCC initiation
will be reduced, and PWSCC CGR may be reduced
in thicker cross-section components, depending upon
other factors such as the stress intensity factor of the
specimen. Andresen et al.11 concluded that crack
growth mitigation by adding Zn requires further
study, although two of four tests show a decrease
in growth rate of >3Â. Molander et al.14 also found
that the effect of zinc on CGR was minor. Hence,
more work is needed before making definitive conclusions from laboratory studies regarding the benefit
of zinc in mitigating CGR.

SG tube nondestructive examination (NDE) data
from eight plants injecting zinc indicated reduction
in the incidence of PWSCC by a factor of 2–10.9 An
example from a 2-unit PWR showing the effect of
zinc on SG tubing over successive cycles is given in
Figure 17.
The largest effect of zinc appears to be on initiation of cracking, with a smaller effect on CGR, with
the data indicating a factor of 2–10 reduction for
initiation and about a factor of 1.5 reduction in CGR,
consistent with the extensive laboratory work,11

Application of zinc in world PWRs

Number of plants and percentage
of PWR injecting zinc

50
45

Percent of PWRs injecting

Number of units injecting

40
35
30
25
20
15
10

5
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Year

Figure 16 Application of zinc injection in pressurized water reactors worldwide.


Water Chemistry Control in LWRs

31

140
X indicates last refueling outage
before
start of zinc injection

Number of new tubes affected

120
100
80

Unit 1
Unit 2
60
40
20
0

X−3

X−2

X−1

X+1
X
Refueling outage

X+2

X+3

X+4

Figure 17 Effect of zinc on steam generator tube degradation at a US pressurized water reactor.

70
60
Number of plants

indicating that zinc inhibits mainly by delaying the
initiation of PWSCC. However, the SG NDE data
also showed that zinc reduced the rate of crack propagation (depth) by 17–60%.
These results are consistent with initial laboratory
test data indicating that zinc reduced crack propagation by a factor of approximately 3 at low stress
intensities, but had no effect at higher stress intensities. In addition, the lack of cracking in the Farley
PWR pressure vessel head penetrations (exposed to
zinc for over 12 years), compared to PWSCC indications in similar pressure vessel heads in other plants,

suggests that zinc addition is beneficial for Alloy 600
(and possibly Alloy 82/182) thick-section components under PWR primary service conditions.
Recent work has studied the influence of dissolved
hydrogen on PWSCC. In the early days of PWR
operation, the lower limit on hydrogen was set at
25 ml kgÀ1, to provide adequate margin against radiolysis and heavy crud formation. Plant tests in France
showed that this limit was excessively conservative
and that less than 10 ml kgÀ1 would be satisfactory,
provided good control of oxygen was maintained in
makeup water. Several workers have found that the
maximum in PWSCC CGR occurs close to the ECP
corresponding to the Ni/NiO thermodynamic equilibrium condition.15 Although this potential is unaffected by lithium/boron/pH (consistent with the fact
that these do not greatly influence PWSCC over the
range of practical relevance), the equilibrium potential is significantly affected by the dissolved hydrogen

50
40
30
20
10
0
25–30
30–35
35–40
40–45
45–50
Cycle average hydrogen concentration (cm3 kg−1)

Figure 18 US plant data for dissolved hydrogen.

concentration. Andresen et al. found that the peak in
SCC growth rate versus H2 fugacity was temperature
dependent, but generally fell within the hydrogen
concentration range used in PWRs. This provides
an opportunity for mitigation, by perhaps a factor of
2 in Alloy 600 and a factor of 5 in Alloys 182, 82,
and X750, as the median value of the dissolved
hydrogen concentration for US plants is approximately 35 ml kgÀ1.
US PWRs currently operate within dissolved
hydrogen within the recommended 25–50 ml kgÀ1
range, with the majority in the 30–40 ml kgÀ1 range,
but none with more than 44 ml kgÀ1 (Figure 18).
The lower limit is set conservatively to provide an
operating margin over the level of hydrogen required


32

Water Chemistry Control in LWRs

to suppress water radiolysis in the reactor core. Somewhat lower concentrations are used in other countries.
The dissolved hydrogen concentrations corresponding to the peak CGR for a typical range of
PWR primary operating temperatures are 4.3 ml kgÀ1
at 290  C, 10.4 ml kgÀ1 at 325  C, and 16.5 ml kgÀ1
at 343  C.15 Andresen et al.11 published Figure 19,
which indicates the proposed factors of improvement
on changing from an initial hydrogen concentration
of 25 ml kgÀ1. It can be seen that raising the hydrogen provides benefit, but lowering it is detrimental
below 330  C.

In response to the data showing the benefit of
increasing hydrogen in reducing CGR, the US industry program in progress focuses on the extent to
which dissolved hydrogen can be increased without
adverse consequences to other parts of the system.
Other countries, including Japan, are also investigating lowering hydrogen, because laboratory data
suggest that the initiation of cracking is delayed at
lower hydrogen concentrations. This is depicted in
Figure 20, as discussed by Molander.14
The lower line in this figure shows the time to
initiate cracking, based on laboratory tests using

Factor of improvement from H2

4.0
Based on Alloy 182,
a current H2 level of 25 cm3 kg−1

3.5

70 cm3 kg-1 H2

25

3.0

45 cm3 kg-1 H2

25

2.5

2.0
1.5

Good
1.0
25
0.5
0.0
270

280

290

1 cm3 kg-1 H2

300
310
320
Temperature (ЊC)

4 cm3 kg-1 H2

25

330

340

Bad

350

Figure 19 Effect of dissolved H2 on primary water stress corrosion cracking crack growth rate at different temperatures.

ml H2/kg H2O (330 ˚C)
10

15

20

25

30

Crack initiation time (h)

Jenssen data on Alloy 600

35

1E−07

Growth
8E−08

20 000

6E−08

15 000

4E−08

10 000
Initiation

2E−08

5000

0
0

5

10
Hydrogen activity (kPa)

15

Crack growth rate (mm s–1)

5
25 000

0E+00
20

Figure 20 Dependencies between the dissolved hydrogen content in pressurized water reactor primary coolant on the
crack initiation time observed on initially smooth surfaces and on the crack propagation rate.


Water Chemistry Control in LWRs

reverse U-bend specimens, whereas the upper line
shows crack growth data over a similar concentration
range. Thus, the lowering of hydrogen appears feasible. However, the relative importance of crack initiation
and crack propagation is very dependent on material
and plant conditions. In the United States, concern
about increased crack propagation at low hydrogen
and low temperatures, as shown in Figure 19, has
resulted in moving to higher hydrogen being preferred
to the alternative of reducing hydrogen. Several factors
combine to make higher H2 the preferred way to mitigate SCC, including the importance of bottom-head
penetrations (which are exposed to $290  C water)
and the recent observation that the CGR in coldworked Alloy 600 is not mitigated at low H2.11
The preferred strategy in the United States is
to gradually increase hydrogen to the upper end of
the existing range, with the potential to move higher
(say to 60 ml kgÀ1) when the ongoing qualification
work is completed. This will include evaluation of
the effects of dissolved hydrogen on radiation fields
and fuel performance, although any such effects are
expected to be minimal.16
5.02.3.3

PWR Radiation Field Control

Corrosion products released from out-of-core materials (primarily SG tubing) deposit on the fuel and
become activated, are released back into the coolant,
and may be deposited on out-of-core surfaces. Both
soluble and insoluble species may be involved, with the
latter tending to deposit in stagnate areas (‘crud traps’).
In addition to the chemistry items discussed later in
this section, it must be stressed that other factors are
important to the goal of reducing radiation fields. In
particular, the success of the later German-designed
plants in eliminating cobalt sources in hardfacing alloys,
thereby achieving very low radiation fields, demonstrates the benefits of cobalt source reduction. With
many plants replacing SGs, a correlation between
recontamination rates and surface finish of the new
SG tubing has been noted by Hussey et al.17
Typical PWR fuel cycles start with a relatively high boric acid concentration, which gradually
reduces to zero at the end of the cycle. Lithium
hydroxide is added to maintain an approximately
constant pH. As the duration of fuel cycles increased,
more boric acid was required at the start of cycle,
which in turn necessitated increased LiOH to maintain the desired pH (Figure 21).
As mentioned earlier, radiation field buildup can
be controlled by minimizing corrosion product

33

transport and activation. Initially, coordination of
lithium hydroxide with boron to maintain a constant
at-temperature pH of 6.9 was recommended, based
on the minimum solubility of magnetite. In fact, the
prime constituent of the crud turned out to be nickel

ferrite, requiring a pH of $7.4 for minimum solubility. Fruzzetti et al.15 have recently reviewed the data
on elevated pH, which provides a number of benefits
including decreased general corrosion (and thus
reduced corrosion product transport to the core).
Field-tests of pHs greater than 6.9 confirmed that
radiation fields were lower. Although no adverse
effects were observed on the fuel, many plants were
slow to abandon a 2.2 ppm limit, established to avoid
excessive zircaloy corrosion. However, there were
indications of heavier crud formation after long periods operating below pH 6.9, and as fuel concerns
relaxed, a gradual move toward a maximum of 3 ppm
lithium resulted. Moreover, pHs in the range 7.1–7.2
became more popular in the late 1990s, with 7.3–7.4
eventually gaining favor.
Figure 22 shows the maximum lithium concentrations reported by US PWRs in recent years. It can
be seen that 95% are now using greater than 3 ppm
at full power: a significant change from earlier in
the decade.
A demonstration of elevated Lithium/pH is in
progress at Comanche Peak PWR.18 The goal was to
reduce radiation fields and reduce susceptibility to the
Axial Offset Anomaly (AOA) by reducing crud
buildup. This test involved increasing the primary
system pH from 7.1/7.2 to 7.3 and then two cycles at
7.4. No significant adverse trends have been noted,
either in the area of chemistry or core performance.
Radiation fields measured have shown a modest but
continued improvement. On the basis of the positive
trends and absence of any negative effects, Comanche
Peak has established elevated constant pHTave 7.4 as

the primary chemistry regime for both units.
Without the increases in pH/lithium that have
taken place, radiation fields would have been
expected to increase significantly for longer fuel
cycles. The increase in boiling in localized regions
of the core (called subcooled nucleate boiling)
in PWRs resulting from power uprating has resulted
in higher crud buildup on the upper fuel surfaces, and
there is growing evidence from US PWRs that radiation fields are indeed higher for the highest rated cores.
Enriched boric acid (EBA), that is boric acid
enriched with B-10, enables a given pH to be
achieved with less lithium hydroxide, as the required
concentration of B-10 can be obtained with less total


34

Water Chemistry Control in LWRs

Constant pH 7.2
6

Lithium
‘Li high limit’
‘Li low limit’

5
Li target = 6.0 E−7 B2 + 0.0023B + 0.4413

Lithium (ppm)

4
3.5 ppm limit
3
2.2 ppm limit
2

Start of 18-month
cycle

1

Start of 12-month cycle

0

20

80

140

200

260

320

380

440

500

560

620

680

740

800

860

920

980

1040

1100

1160

1220

1280

1340

1400

1460

1520

1580

1640

1700

0

Boron (ppm)
Figure 21 Lithium concentrations required to maintain pH 7.2 for different fuel cycle lengths.

boric acid. EBA is used at several plants in Europe,
typically to increase shutdown margin when using
mixed oxide fuel (MOX), but has not been applied
to date in the United States. However, consideration
is being given to using EBA at some plants that will
use MOX fuel in the future. Despite the transition to
the use of EBA in operating plants, designing for it in
new plants is recommended.19
As discussed earlier, the motivation for the initial
applications of zinc in most US PWRs was to control
PWSCC of SG tubing. However, German-designed

PWRs and a few US plants used $5 ppb depleted
zinc for radiation control, mostly with depleted zinc
to avoid zinc-65 formation. A recent paper ‘‘Understanding the zinc behavior in PWR primary coolant: a
comparison between French and German experience’’
by Tigeras et al.20 provides a European perspective on
this topic.
This paper concludes that ‘zinc injection seems to
present the most positive and clearest results: in all
the units injecting zinc, a dose rate reduction has
been detected after a certain period of exposure
without leading to any negative impact on plant

systems, components, and operation.’ Thus ‘zinc
injection should be considered as a strategy with
benefits in short, medium, and long term. Its application as soon as possible in the life of nuclear power
plants and especially before SG replacement and fuel
cycles modifications seems to be an excellent decision to contribute to ensuring the passivation process
of new components, the fuel performance, the full
power operation of the units, and the long life of
materials and components.’
Figure 23 shows the effect of zinc in reducing
radiation dose rates at several plants. It can be seen
that the reduction factor approximately correlates
with the cumulative zinc exposure in ppb months
(the product of the average zinc concentration and
the duration of zinc addition). As little as 5 ppb
zinc has been shown to reduce radiation fields by
35–50% at operating plants, based on zinc exposures
of !700 ppb months. There is relatively little difference between plants with Alloy 600/690 SG tubing
and those with Alloy 800 tubing, but plants using

depleted zinc show greater benefit than those using
natural zinc, as shown in the figure.


Water Chemistry Control in LWRs

35

Percentage of units within range

70

<3 ppm
3.0−3.5 ppm
>3.5 ppm

60
50
40
30
20
10
0

2000

2001

2002

2003

2004

2005

2006

2007

EOC year
Figure 22 Maximum reported coolant lithium (full power) at US pressurized water reactors.

Cumulative dose rate reduction fraction

1.2
Alloy 800 w/depleted zinc
Alloy 600 and 690 w/depleted zinc
Alloy 600 and 690 w/natural zinc
Log Alloy 800 plants
Log Alloy 600 and 690 w/depleted zinc
Log Alloy 600 and 690 w/natural zinc

1

0.8

0.6

0.4

0.2

0
0

200

400

600
800 1000 1200 1400 1600
Cumulative zinc exposure (ppb months)

1800

2000

Figure 23 Effect of zinc injection on radiation dose rates.

5.02.3.4

Fuel Performance

With progressive uprating of fuel duty, the margin to
tolerate crud has been reduced and additional care
has to be taken in specifying the water chemistry to
avoid undesirable fuel performance issues. Figure 24
shows the root causes of PWR fuel failures since
2000, including failures from pellet–clad interaction

causing SCC, fabrication defects, debris, grid fretting, and crud/corrosion. In contrast to the BWR

situation, shown in Figure 11, very few failures in
recent years have been attributed to crud/corrosion
(the exceptions to this comment are discussed in a
following section).
A phenomenon called axial offset (AO) has caused
concern over the past 10 years.21 AO is a measure of
the relative power produced in the upper and lower
parts of the core and is normally expressed as a
percent, with a positive percent indicating that


36

Water Chemistry Control in LWRs

Number of failed assemblies

120
Unknown
Debris
Crud/corrosion

100

Fabrication
PCI-SCC
Grid fretting

80
60
40
20
0
2000

2001

2002

2003

2004

2005

2006

2007

EOC year
Figure 24 US pressurized water reactor fuel failures by mechanism.

more power is produced in the upper part of the core.
AOA occurs when boron concentrates in corrosion
product deposits (crud) on the upper spans of fuel
assemblies undergoing subcooled nucleate boiling,
causing a reduction in neutron flux. AOA has affected
at least 20 PWRs in the United States, as well as

several in other countries.
Clearly, fuel crud is involved in the AO phenomenon, and water chemistry effects must be considered
in controlling AO. Besides their axial asymmetry, the
composition of fuel deposits in boiling cores is different from nonboiling fuel. The nickel-rich deposits on
boiling cores tend to be removed much less effectively
by conventional chemistry shutdown evolutions than
the nickel-ferrite deposits on nonboiling cores. Alternative methods are therefore required for removing
corrosion product deposits from reload fuel from highduty cores, including ultrasonic fuel cleaning.
An important difference exists between plants
with Alloy 600 or 690 SG tubing and those (such as
German-designed plants) with Alloy 800 tubing.
The latter have a much lower proportion of nickel
in fuel crud and have not experienced the AO
phenomenon.22
Early work showed that lithium increased zircaloy oxidation rates, although the adverse effects
were reduced in the presence of boric acid. As a
result, a limit of 2.2 ppm lithium was generally
imposed to reduce zircaloy corrosion, although
excessive crud formation at low pHs was likely to
be more detrimental to the cladding than higher
lithium concentrations, especially as the resistance

to corrosion of zircaloy improved. This was confirmed by one of the few failures in recent years
that was uniquely attributed to crud buildup. In this
example, a move to a longer fuel cycle necessitated
increasing the boron concentration at start of cycle;
however, the 2.2 ppm lithium limit was retained,
resulting in the pH being well below 6.9 for the
initial period of the cycle. This in turn caused
heavy crud formation, to which subsequent fuel failures were attributed.

The move in the past ten years to greater fuel duty,
with operation of fuel at higher temperatures (with
localized subcooled nucleate boiling), has caused
crud-related problems to reappear, particularly the
localized flux depression as a result of buildup of
boron-containing crud, which were discussed earlier.
This in turn has renewed interest in elevated pH/
lithium to minimize corrosion product transport, the
use of EBA and the more immediate mitigation that
can be obtained from fuel cleaning.
Fuel performance is always a concern with changes
in water chemistry, such as zinc injection. On the basis
of current experience, the impact appears to be minimal for the majority of plants, but insufficient data
exist for plants with the highest fuel duties to allow
application without postexposure fuel inspections.
Data from US plants suggest little or no fuel concerns
for coolant zinc levels up to 40 ppb for plants with
less-highly rated cores. Extended experience at these
plants, over at least 10 years of operation, indicates no
adverse effects on fuel at zinc concentrations from
15 to 25 ppb. However, there have been no data


Water Chemistry Control in LWRs

available until recently for higher zinc concentrations
in higher duty cores where significant subcooled
nucleate boiling occurs on the fuel clad surface.23
Perkins et al.24 comment that fuel performance must
be considered prior to injecting zinc and additional

monitoring and fuel surveillances to understand and
evaluate the impact and the role of zinc may be
required in some circumstances.

5.02.4 PWR Secondary System Water
Chemistry Experience
5.02.4.1 Evolution of PWR Secondary
Chemistry Strategies
The objectives of PWR secondary water chemistry
control are to maximize secondary system integrity
and reliability by minimizing impurity ingress and
transport, minimizing SG fouling, and minimizing corrosion damage of SG tubes. Since secondary side
corrosion damage of SG tubes is primarily caused by
impurities in boiling regions, where high concentrations of impurities occur in occluded regions of
the SG formed by corrosion product deposits, new
approaches are continually sought to control corrosion
product transport to and fouling within the SGs.25
PWRs have experienced IGA on both the primary
and secondary sides of the Alloy 600 SG tubing, which
has been a major contributing cause of the replacement of most of the SGs with mill-annealed tubing,
not only in the United States but internationally.
Figure 25 illustrates the various corrosion processes
found in different locations in a recirculating SG.26
PWR secondary system water chemistry has
evolved through many changes over the years, largely
in response to emerging technical issues associated
with this degradation of structural materials in SGs.
In the early days of PWR operation, wastage became
a problem in the secondary side of PWR SGs, resulting in a switch from the use of sodium phosphate
inhibitor to all-volatile treatment (AVT) using

ammonia, which in turn brought about the denting
phenomenon. Tighter control of impurities, oxidizing potential, and pH were necessary to mitigate
the denting problem. Despite continued chemistry
improvements, many plants have had to replace
SGs of earlier designs (e.g., those tubed with Alloy
600MA), as shown in Figure 26.
Newer generation SGs are performing well,
although there remain concerns about the adverse
effects of lead impurity, causing Pb-assisted stress
corrosion cracking (PbSCC), which is discussed later.

37

Lead has been observed in various flow streams (final
feedwater, heater drains, etc.) in the secondary systems
of PWRs. Lead is detected at some concentration in
nearly all deposit analyses (SG and other locations).
Lead is present in trace concentrations in secondary
system materials of construction, as well as in chemical
additives such as hydrazine.15
Figure 27 shows the worldwide causes of SG
repairs through 2004. It can be seen that IGA is
currently the most prevalent form of degradation.
Figure 28 compares the behavior of three types of
SG tubing, Alloy 600MA (mill-annealed material used
in early plants, Alloy 600TT (thermally treated material used in later plants), and Alloy 690TT (an
improved alloy used in most replacement SGs). This
diagram is taken from the 2008 PWR Secondary Water
Chemistry Guidelines,27 which contains a much more
detailed account of corrosion processes. 600TT has

reduced susceptibility under mildly oxidizing highalkaline conditions, that is, SCC is not observed until
higher pH than for 600MA, and 600TT has approximately the same susceptibility as 600MA under acidic
conditions. 690TT is indicated as having a still smaller
region of susceptibility in the high-alkaline region and
as having no susceptibility in the acid region except
under highly oxidizing conditions that are unlikely to
occur in plants. However, other work indicates that
SCC can occur in 690TT at an acidic pH, especially
if lead is present. Also, SCC occurs in both 600MA and
600TT in the mid pH region if lead is present.
In the 1990s, improved pH control using amines
became a regular practice, and fine-tuning, including
using mixtures of different amines to control pH
throughout the circuit and coordination with resin
utilization, continues today. Hydrazine is used to
remove oxygen from the system. Hydrazine levels
have continually been reviewed and ‘optimized,’ with
due regard to any impact on FAC in secondary systems, as FAC rates increase at very low oxygen concentrations. Molar ratio control (MRC) describes a
control strategy that adjusts the bulk water chemistry,
generally sodium and chloride, such that the solution
that is developed in the flow-occluded region is targeted to be near neutral. MRC can involve the addition of chloride ions to ‘balance’ the cations that
cannot be reduced via source term reduction programs. MRC was widely practiced to minimize SCC
concerns, but has not been actively employed at plants
replacing to SGs tubed with Alloy 690TT. With
more plants replacing their SGs, less plants are adopting the MRC program. Only ten plants were doing
MRC in 2007,28 and they are all with original SGs with


38

Water Chemistry Control in LWRs

U-bend cracks
(PWSCC)

Fatigue

Free span
ODSCC
IGA

ODSCC

PWSCC

Expansion
transition

PWSCC

PWSCC or
ODSCC

ODSCC
Denting

Fretting, wear, corrosion,
thinning

Tubesheet

Pitting
IGA

Expansion
transition

Tubesheet

ODSCC
Sludge

Tubesheet

PWSCC tube-end
cracking

Tubesheet

Figure 25 Corrosion processes in recirculating steam generators, showing primary water stress corrosion cracking and
outside diameter stress corrosion cracking on the secondary side.


Water Chemistry Control in LWRs

39

140
Operating plants
Plants w/replacement SGs
120

134

84
88

81

134
134

134
72

64
67

59

52

45
51

37

22
29

17

10

7

7

7

7

3
5
7

2

78

134

134

132

131
132

134
131

131

132

132
133

134

12

132

134

12

133

133

130

121

63
68

55

46

1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000

2001
2002
2003
2004
2005
2006
2007

1

22

20

31

41

40

0

11

93
99

60

108

115

127

80

79

Number of plants

100

Year
Figure 26 Steam generator replacement status worldwide.

100
90
80

Percent

70
60
50
40
30
20
10
1973
1974

1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004

0
Year
IGA
Impingement Pitting
Other
Wear
Thinning

Fatigue
Unknown
SCC
Preventive

Figure 27 Worldwide causes of steam generator tube repair.

600MA and 600TT tubing. Currently, no plants with
replaced SGs are believed to be using MRC.
Titanium-based inhibitors to minimize corrosion are also employed at some plants. Boric acid

treatment (BAT) involves the addition of boric acid
to feedwater. Such approaches are worthy of consideration, on the basis of plant-specific degradation mechanisms, operational considerations, and


40

Water Chemistry Control in LWRs

1.0

TT690

Potential (V vs. Ec )

0.8
Some tests indicate that 690 TT may be
susceptible in the low pH region, especially if lead is
690 TT U-bend cracked
present
in near neutral AVT
with lead and oxidizing sludge
TT690
TT600
TT600
MA600

0.6

0.4

MA600
0.2
600 MA and 600 TT can be susceptible
in mid pH range if lead or reduced
sulfur is present.

0

2

3

4

5

6
7
8
pH 300 °C (572 °F)

9

10

11

12

Figure 28 Corrosion mode diagram for Alloys 600MA, 600TT and 690TT (based on Constant Extension Rate Tensile
Tests at 300  C), showing regions where materials are susceptible to attack.

interactions. The most recent developments are
aimed at reducing deposit buildup in crevices,
including the use of dispersants, such as polyacrylic
acid (PAA), that is discussed in more detail later.
The historical trends in PWR secondary chemistry are shown in Figure 29.
5.02.4.2 Chemistry Effects on Materials
Degradation of SGs
Corrosion of SG tubes has been the major issue

affecting selection of secondary water chemistry
parameters. However, corrosion and FAC of SG
internals and other secondary system components
are also important concerns.
Corrosion of SG tube materials is mainly affected
by the following water chemistry related factors, in
addition to nonwater chemistry factors such as material susceptibility, temperature, and stress:
 pH – Corrosion of several different types, including IGA/SCC and pitting, are strongly affected
by the local pH. High pH (caustic conditions)
and low pH (acidic conditions) accelerate the
rates of IGA/SCC.
 ECP – The ECP is a measure of the strength of the
oxidizing or reducing conditions present at the

metal surface. The rate of corrosion processes are
strongly affected by the ECP. Secondary side SCC in
tube alloys tends to be accelerated by increases in
ECP, that is, by the presence of oxidizing conditions.
 Specific species – Some impurity species accelerate
corrosion of tubing alloys as a result of their effects
on pH and ECP. In addition, lead and reduced
sulfur species (e.g., sulfides) appear to interfere
with formation of protective oxide films on the
tube metal surfaces, and thereby increase risks of
IGA/SCC, independent of influences on pH or
potential. Similarly, chlorides tend to increase the
probability of pitting.
These factors have been most thoroughly explored
for mill-annealed Alloy 600 (600MA). As discussed
in Chapter 5.04, Corrosion and Stress Corrosion

Cracking of Ni-Base Alloys, tests indicate that the
other tubing alloys, that is, stress-relieved Alloy
600 (600SR), thermally treated Alloy 600 (600TT),
nuclear grade Alloy 800 (800NG), and thermally
treated Alloy 690 (690TT), exhibit similar tendencies, but have increased resistance to corrosive attack,
in the order listed, with 690TT having the highest
resistance. Laboratory tests and plant experience indicate that 690TT has very high resistance to IGA/SCC
on the outside diameter on tubing (OD IGA/SCC) in


Water Chemistry Control in LWRs

41

Pb remediation
Dispersants
Titanium
Molar ratio control
MPA, DMA
ETA chemistry
Morpholine chemistry
Boric acid addition
EPRI water chemistry guidelines
Ammonia chemistry
Phosphate
1975

1980

1985

1990

1995

2000

2005

Figure 29 Evolution of water chemistry for pressurized water reactor secondary systems.

normally expected crevice conditions, but OD IGA/
SCC could possibly occur as a result of upsets or as a
result of long-term fouling and accumulation of aggressive species in deposit-formed crevices. Alloy 800NG
also has high resistance to OD IGA/SCC, but laboratory tests indicate that it is about twice as susceptible as
Alloy 690TT, and it has experienced limited amounts
of IGA/SCC in plants, while no operation-related
corrosion of 690TT has been reported. Laboratory
tests and some plant experience indicate that 600TT
is significantly more resistant than 600MA but less
resistant than 800NG and 690TT.
Water chemistry selected to protect SG tubes
appears to be satisfactory for most balance-of-plant
(BOP) components such as turbines. The main corrosion concerns in the BOP that affect secondary
system water chemistry are FAC of carbon steel piping, tubing, and heat exchanger internals and shells,
and ‘ammonia’ attack of copper and copper alloy
tubes. In addition, FAC has also affected some recirculating SG internal components (e.g., feedrings,
swirl vanes). FAC is mainly influenced by the
at-temperature pH and oxygen content around the
secondary system. ‘Ammonia’ attack of copper alloys

is mainly influenced by the concentrations of ammonia and oxygen at the copper alloy locations, but is
also accelerated by increases in concentrations and
pH associated with other amines, although not as
strongly as by increases in ammonia.
Once-through steam generators (OTSGs) have
different thermal hydraulics and (in original SGs)

tube materials than recirculating steam generators
(RSGs). These differences have led to OTSGs
having somewhat different tube corrosion experience than RSGs of the same vintage. For the most
part, OTSGs have experienced somewhat lower
rates of tube degradation. However, significant
IGA/IGSCC has been detected in the upper bundle
free spans of several units, especially at scratches,
and SG replacement has been performed or is
planned at all units.
The locations in SGs that are most affected by
IGA/IGSCC are those where free circulation of secondary water is impeded by the local geometry, for
example, in crevices formed by tube support plates or
by sludge piles that can accumulate on the tube sheet.
Impurities in the secondary water can concentrate in
these locations by boiling and evaporation in a process called ‘hideout.’
The key issue influencing water chemistry regimes
in PWR secondary system is to minimize SG degradation by controlling sludge buildup, reducing (and
balancing, e.g., MRC) the concentration of impurities
(i.e., sodium, chloride and sulfate) in deposits at the
tube-tubesheet and tube-tube support plate interfaces. The use of advanced amines to control pH has
increased significantly in the past few years, as discussed in a following section. Figure 30 shows the
main approaches used in typical chemistry control
strategies.

Impurities are removed from SGs by blowdown of
the coolant. Over the past 20 years or so, average