See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228503085

Behaviour of Copper in Generator Stator Cooling-Water Systems

Article · January 2009

CITATIONS READS

11 6,687

2 authors:

Robert Svoboda Donald Alan Palmer

47 PUBLICATIONS   108 CITATIONS   
Oak Ridge National Laboratory
285 PUBLICATIONS   8,885 CITATIONS   
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Chemistry in steam turbines View project

Chemistry in water cooled generators View project

All content following this page was uploaded by Donald Alan Palmer on 05 June 2014.

The user has requested enhancement of the downloaded file.

 P R E P R I N T – ICPWS XV
Berlin, September 8–11, 2008

Behaviour of Copper in Generator Stator Cooling-Water Systems

Robert Svoboda a and Donald A. Palmer b

a
Alstom Power, Brown Boveri Str.7, CH-5401 Baden, Switzerland
b
Oak Ridge National Laboratory, Box 2008, Building 4500S, Oak Ridge, TN 37831-6110, USA
Email: [email protected]

Water-cooled generators with hollow copper strands frequently suffer from deposition of copper oxides
that clog them and thus impair cooling water flow. Solubility is one of the factors governing the release and
the re-deposition of copper oxides. Results presented in this paper indicate that under generator cooling water
conditions the solubility of copper oxides is dependent on pH and also, to a lesser degree, on temperature. At
neutral pH, the solubility of copper oxides increases with increasing temperature. The dependence is stronger
for Cu2O and small for CuO. With acidic pH, the solubility of CuO may even slightly decrease with
temperature. The apparently contradictory information in the literature - whether the solubility decreases or
increases with increasing temperature - is probably a result of differing test conditions, or a differing mix of
copper oxides, or to experimental differences.

Introduction
Copper itself is stable in pure water. If oxygen
Water-cooled generators with hollow copper is added, copper(I) oxide and / or copper(II) oxide
strands ("hollow conductors") frequently suffer are formed, depending on the electrochemical
from deposition of copper oxides that clog them potential (ECP) [2][3]. Laboratory-scale
and thus impair cooling water flow, Figure 1. experiments at ORNL have shown that ingress of
Various water treatment regimes have been oxygenated water to a column of clean copper
developed in order to mitigate this problem [1]. The beads at 100 oC, can lead to copper release well in
aim of these water treatment regimes is to control excess of that solubility limits of either oxide so
formation, release, migration, and re-deposition of that presumably the kinetics of copper metal
copper oxides, either individually or in combination, oxidation and release can exceed that of copper
Figure 2 and reference [2]. oxide film formation [4]. Nevertheless, the release
and re-deposition of these oxides play a pivotal role
in the clogging of hollow conductors.

Figure 2: Model for copper oxide deposits

1 formation (oxidation)
2 release
Figure 1: Copper oxide deposits at the hollow 3 migration
conductor inlet 4 re-deposition
 Figure 3:
Summary of copper release
rates in function of oxygen
level and pH. Operating
areas of various chemistry
regimes [2]

There is a wealth of information published on

parameters influencing release and deposition [1] to
[13], some of which will be discussed more closely
in this paper. Figure 3 summarizes the influence of
pH and oxygen concentration, and also shows their
relation to the current water treatment regimes.
This paper is following up reference [2], which
noted that international literature presented
conflicting information on the temperature
dependence of the solubility of copper oxides.

Morphology of hollow conductor deposits Figure 4: Oxides from oxidation of a fresh copper
surface in laboratory loop: at pH=8 (left) and pH=7
Laboratory tests with clean copper surfaces (center). Pile of oxide flakes found in a generator
under simulated stator operating conditions yielded stator (right). Scale: the horizontal width of each
crystalline oxide layers. picture corresponds to 5 μm.
The morphology of the oxides was influenced
by the system water chemistry. Figure 4 gives
examples of a specimen with dense polyhedral
oxides, typical for high-oxygen or alkaline water
chemistry, and another one with flower-shaped
needles, typical for low-oxygen or neutral water
chemistry. Coupon samples in a generator stator
cooling system verified these results.
With long-term generator operation, the picture
becomes more complex. Deposits can be of either
predominantly crystallized growth, Figure 1, or an
agglomeration of broken-off particles, or a
Figure 5: Oxide deposit in a used hollow conductor
combination of both, Figures 4 and 5.
taken from a generator. Flow is from right to left.
The oxides from a generator stator are always a
mixture between Cu2O and CuO, with Cu2O first step of excessive deposition is ripple-like
dominating in low-oxygen systems, and CuO growth, Figure 5. An explanation of this
dominating in high-oxygen systems. dependence on hydrodynamics could be the
There is evidence that an accumulation of increased mass transport from the bulk liquid to the
deposit growth occurs preferentially in areas of copper surface, enhancing crystallographic growth.
increased turbulence, that is: at the inlets, outlets of As the water velocity in hollow conductors is
hollow conductors and inside its length also on moderate (usually 0.5 - 2 m/s), Flow Accelerated
bends and Roebel transpositions. Very often, the

2
Figure 6: Compact oxide plug in a used hollow Figure 7: Pourbaix diagram and potential of
conductor taken from a generator. passivated (black CuO layer) and active (freshly
pickled) copper specimens [12]

Corrosion is improbable under neutral and alkaline Temperature effect on the copper release rate
conditions. This is supported by field experience.
With aging, the oxides become more compact, Literature data
Figure 6, possibly by crystallographic deposition in It is a well known experience from fossil plant
porous deposits. Often, the aged deposits also chemistry that copper-alloy tubed feedwater heaters
contain important fractions of metallic copper and release copper, and the higher the temperature level
become hard to remove. The occurrence of metallic is, the higher is the copper release [14]. It has
copper may be the result of deposition of possible however to be taken into account that these plants
colloidal Cu particles. It is also possible that copper usually operate with ammonia and possibly also
oxides are reduced to metallic copper. The with hydrazine which makes copper chemistry
dissolved hydrogen in stator water alone however more complex than it is in a water cooled generator.
does not provide sufficient electrochemical driving It is therefore of special interest to follow the
force, Figure 7 [12]. Possibly, disproportion of investigations done by Brush and Pearl in the late
Cu2O into Cu and CuO may take place. It should be 1960's for BWR cycle chemistry with neutral
mentioned that metallic copper deposits are also feedwater, Figure 8 [15][16]. The findings are
found in generator cooling systems outside of the similar to those from fossil plants: copper release
generator, like on stainless steel pump wheels or increases with rising temperature.
hard metal pump seals. These results are however only indicative,
because these investigations related to various
Effect of temperature on copper oxide deposits copper alloys, but in generators pure copper is used.

Research by Hwang, Park, and Rhee asserts that

hollow conductor plugging is mainly caused by the
excessive release of CuO and Cu2O particles. This
release was found to be dependent on the ECP [3].
Other researchers consider that variations in the
solubility of copper oxides in water cause local
deposition and plugging. Here, the temperature
dependence of the solubility is of relevance,
because the copper strands are subject to a
temperature increase along the water path through
the generator. If, for example, solubility decreases
with increasing temperature, copper oxide is
dissolved at the colder inlet and re-deposited
towards the warmer outlet of the hollow conductor Figure 8: Steady state copper release in neutral
[13]. Literature data however give apparently feedwater at 5 ft/s containing (left) low oxygen (3 μg/kg
contradictory information on such a trend of /kg O2 and 100-200 μkg /kg H2), and (right) elevated
temperature dependence. This shall be explored in oxygen (200 μg/kg /kg O2 and 25 μkg /kg H2) [15][16].
some detail. „ Alloy 400, ▲Admirality brass,
† CuNi 70/30, ‹ CuNiFe 70/30, ◊ CuNi 90/10,

3
Laboratory tests on copper release rate 100
Laboratory investigations were made on the
parameters for copper release rates in BBC
technology generators. Some data have been

Cu release rate (gCu/m2.d)

previously published [2][12]. The laboratory test
loop was equipped with features that allowed
control of the water velocity, temperature, oxygen,
pH, and chemical additives. The tests were made on 10
pure copper.
It is however problematic to measure pH O2 20 ppb
directly in a closed test loop with high purity water. O2 200 ppb
It was therefore measured indirectly via
Cu release rate at
conductivity. Spot checks were made with a single-
pH25°C = 6.5-7.0, v = 2.2 m/s
electrode device that had been conditioned in a
neutral salt solution. Giving allowance for these
1
uncertainties, a pH range of 6.5-7.0 is given for the
15 20 25 30 35 40 45 50 55 60
absolute value of pH. The actual pH was then kept
Tem perature (°C)
constant throughout the tests. The results, Figure 9,
indicate that the release rate increases with Figure 9: Copper release rate in a laboratory test loop.
increasing temperature.

Significance of the copper release rate

The drawback of these considerations however
is that the release rate is only one contributing
factor to oxide deposits. Once the copper is released
it may re-deposit (and that may depend on the
existence of previous deposits that act as nucleation
seeds), but it may also be properly removed by the
stator water mixed bed.

Temperature effect on copper oxide solubility

The solubility of copper oxides has been

extensively studied in order to understand copper
transport in the thermal cycle of power plants.
Reference [14] gives an overview. The majority of
work focuses on cycles treated with alkalizing
agents like ammonia and hydrazine that may also
react with copper oxides. Data from such complex
chemistries are however not readily applicable to
generator stator systems. On the other hand, data
from investigations in neutral, high purity water are
here of direct interest [4] and [14] to [22].
Figure 10: Experimental data for the solubility of copper
While most research indicates that oxide
oxides, in dependence of pH (at temperature) and of
solubility increases with temperature, other research temperature.
point to a reverse relationship [13]. Different test
conditions and the influence of stronger parameters
like the pH may account for this discrepancy. Cu2O and of CuO in dependence of pH and for
temperatures of 50°C and 100°C [4]. Although
100°C is not attained at normal generator cooling
Laboratory tests on copper oxide solubility
conditions, this temperature nevertheless serves as
The laboratory tests at ORNL are described in
an anchor-point for the temperature trend. For
earlier publications [4][17]. A selection of these
reasons of clarity, other temperatures, like 25°C
results was made with regard to conditions in stator
have not been included in the graph.
water systems. Figure 10 presents the solubility of

4
 Summary and Conclusions

Results presented in this paper indicate that the

solubility of copper oxides are dependent on pH
and also, to a lesser degree, on temperature.
At neutral pH, the solubility of copper oxides
increases with increasing temperature. The
dependence is stronger for Cu2O and small for CuO.
With acidic pH, the solubility of CuO may even
slightly decrease with temperature.
The apparently contradictory information in the
literature - whether the solubility decreases or
increases with increasing temperature - is probably
a result of differing test conditions, or a differing
mix of copper oxides, or due to experimental
differences such as: the control and measurement of
pH; the characterization of the copper oxide before
and after the experimental sequence; continuous
control of the redox state of the system; the
attainment of a true equilibrium condition.
The model that oxide deposits take place
because their solubility decreases at the higher
temperature of a stator bar outlet could not be
confirmed: the solubility increases with temperature.
Figure 11: Solubility of Cu2O and of CuO at neutral pH.
Further, oxide deposits do not only occur at the
The vertical lines indicate the neutral pH at 50°C and
100°C, resp. Their intersection points with the respective warm conductor outlets, but also at the colder
solubility curves give the solubilities at pHN. Full lines conductor inlets, Figure 1 and reference [8], which
and round intersection points are for 50°C, dotted lines contradicts this model.
and triangular intersection points for 100°C. Reference [3] indicates that hollow conductor
plugging may be related to the excessive release of
CuO and Cu2O particles, in dependence on the ECP.
Analysis of the test data
Figure 11 analyzes the conditions at neutral pH, Acknowledgements
which is 6.64 at 50°C, and 6.13 at 100°C. It is seen
that at a constant neutral pH, the solubility of both The investigations on copper release rates were
CuO and Cu2O at 100°C are higher than at 50°C, performed by B. Eugster and R.O. Müller in the
but the change is less pronounced for CuO. Alstom laboratories in Baden/Switzerland. The
This positive dependence of solubility on investigations at ORNL were sponsored by EPRI,
temperature is also seen at very slightly acidic and Inc. under direction of R.B. Dooley.
at slightly alkaline pH values, such as encountered
with neutral or alkaline generator water treatment.. References
The trend can however be reversed and
solubility at 100°C may be smaller than at 50°C in [1] Low Temperature Corrosion Problems in
Fossil Power Plants - State of Knowledge
the following cases:
Report, Chapter 16. EPRI report 1004924,
• at acidic pH, which is within the range of December 2003
some stator water treatment regimes
[2] R.Svoboda, H.G.Seipp: Flow restrictions in
• at intermediate alkaline pH (for example with water cooled generator stator coils -prevention,
NaOH, pH25°C ca. 10-11), which is however not diagnosis and removal. Part 1: Behaviour of
any more relevant for generator water copper in water-cooled generator coils.
chemistry Power Plant Chemistry, 6 (2004) 1, pg 7-14
[3] Prevention of Flow Restrictions in Generator
Stator Water Cooling Circuits. EPRI report,
1006684, February 2002

5
 [4] Behavior of Aqueous Electrolytes in Steam [14] B.Syrett: A Review of Corrosion of Copper
Cycles. The Final Report on the Solubility and Alloys Exposed to High Purity Water.
Volatility of Copper(I) and Copper(II) Oxides. International VGB-EPRI-Conference on
EPRI Report, 1011075, September 2004. Interaction of Non-Iron Based Materials with
[5] P.H. Effertz and W. Fichte: Beeinflussung der Water and Steam, Piacenza / IT, June 1996
Kupferkorrosion in hochreinem Wasser [15] E.G.Brush, W.L.Pearl: Corrosion and
(Influencing Copper Corrosion in High Purity Corrosion Product Release in Neutral
Water). Jahrbuch Vom Wasser, 1974. Feedwater. Corrosion 28(1972)4, pg 129-136
[6] H.G.Seipp: Das Korrosionsverhalten von [16] E.G.Brush,W.L.Pearl: Corrosion Behavior of
Kupfer in wassergekühlten Nonferrous Alooy Feedwater Heater Materials
Generatorwicklungen (The corrosion in Neutral Water with Low Oxygen Contents.
behaviour of copper in water cooled generator Corrosion 25(1969)3, pg 99-105
coils). VGB Kraftwerkstechnik 59(1979)3, pg [17] D.A.Palmer, B.Bénézeth, J.M.Simonson: The
245-248 Solubility of Copper Oxides around the
[7] W. Oschmann: Probleme in Kühlsystemen Water-Steam Cycle. 7th International EPRI
Wassergekühlter Generatoren (Problems Conference on Cycle Chemistry in Fossil
Encountered in Generator Water Cooling Plants. Houston / USA, June 2003
Systems). VGB Conference on Power Plant [18] F.J. Pocock and J.F. Stewart: The Solubility
Chemistry, Essen, Germany. Oct 1979. of Copper and Its Oxides in Supercritical
[8] K.Schleithoff and H.W.Emshoff: Optimierung Steam. J. Eng. Power, 33-45 (1963).
der Konditionierung von Generatorkühlwasser [19] B.Hearn, M.R.Hunt, and A.Hayward:
(Optimisation of the Conditioning of Stator Solubility of Cupric Oxide in Pure Subcritical
Cooling Water). VGB Kraftwerkstechnik 70, and Supercritical Water. J. Chem. Eng. Data,
(1990)9, pg 794-798. 14, 442 - 447 (1969).
[9] T.Fritsch, M.Molière, Y.Verdier, P.Calle, [20] L.N. Var’yash: Hydrolysis of Cu(II) at 25-
O.Menet, D.Noel, M.Waciega: Corrosion des 350°C. Geokhim., 1003 - 1013 (1985).
conducteures creux de stators d'alternateurs
[21] L.N. Var’yash: Equilibria in the Cu-Cu2O-
refroidis par eau (Corrosion of Hollow
H2O System at 150-450°C. Geochem. Int., 26,
Conductors of Water Cooled Generator
80 – 90 (1989).
Stators. SFEN Conference on Contribution of
Materials Investigation to the Resolution of [22] S.E.Ziemniak, M.E. Jones and K.E.S.Combs:
Problems Encountered in Pressurized Water Copper(II) Oxide Solubility Behavior in
Reactor, Fontevraud, France, Sep 2 6, 1985. Aqueous Sodium Phosphate Solutions at
Elevated Temperatures. J. Solution Chem., 29,
[10] M.Molière, Y.Verdier, C.Leymoine: Oxidation
179 - 200 (1992).
of Copper in High Purity Water at 70°C:
Application to Electric Generator Operation.
Corrosion Science 30(1990),2/3, pg 183-188
[11] S.LeBon, M.Dordonat, J.C.Gabriel,
D..Vermeeren, J.F.Combes, S.Perrin:
Bouchage des Conducteurs Creux
d'Alternateur - Mecanisme – Etude
Thermohydraulique - Traitments Curatif et
Preventif (Plugging of Generator Hollow
Conductors - Mechanism - Thermohydraulic
Study - Cures and Prevention). 4th SFEN
Conference on Contribution of Materials
Investigation to the Resolution of Problems
Encountered in Pressurized Water Reactor,
Fontevraud, France, Sep 1998.
[12] R.Svoboda, H.Sandmann, H.G.Seipp, C.Liehr:
Water Chemistry in Generator Water Cooling
Systems. International VGB-EPRI-
Conference on Interaction of Non-Iron Based
Materials with Water and Steam, Piacenza / IT,
June 1996
[13] F.Duffeau, D.Tarrant: CIGRE Guide on
Stator Water Chemistry Management. Draft
Version 4, 2008.

View publication stats