You are on page 1of 242

State-of-Knowledge of Copper in

Fossil Plant Cycles


TR-108460

Final Report, September 1997


Effective December 6, 2006, this report has been made publicly available in accordance with
Section 734.3(b)(3) and published in accordance with Section 734.7 of the U.S. Export
Administration Regulations. As a result of this publication, this report is subject to only
copyrightprotection and does not require any license agreement from EPRI. This notice
supersedes theexport control restrictions and any proprietary licensed material notices
embedded in thedocument prior to publication.

Prepared by
R. B. Dooley, EPRI
B. C. Syrett, EPRI
T. H. McCloskey, EPRI
J. Tsou, EPRI
K. J. Shields, Sheppard T. Powell Associates, LLC
D. D. Macdonald, Pennsylvania State University

Prepared for
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, California 94304
EPRI Project Manager
R. B. Dooley
Strategic Research and Development
Strategic Development Activities
for
Fossil Boiler O&M Cost Reduction Target
Generation Group

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES


THIS REPORT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED
OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF
EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM:
(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE
USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS REPORT,
INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT
INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY,
OR (III) THAT THIS REPORT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR
(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY
CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY
OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS REPORT OR ANY INFORMATION,
APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS REPORT.
ORGANIZATION(S) THAT PREPARED THIS REPORT
ELECTRIC POWER RESEARCH INSTITUTE

ORDERING INFORMATION
Requests for copies of this report should be directed to the EPRI Distribution Center, 207 Coggins
Drive, P.O. Box 23205, Pleasant Hill, CA 94523, (510) 934-4212.
Electric Power Research Institute and EPRI are registered service marks of Electric Power Research Institute, Inc.
EPRI. POWERING PROGRESS is a service mark of Electric Power Research Institute, Inc.
Copyright 1997 Electric Power Research Institute, Inc. All rights reserved.

REPORT SUMMARY
Uncontrolled copper transport activity represents a potentially significant source of
performance and reliability loss to fossil plant units with mixed metallurgy feedwater
systems. Recent utility experiences with severe copper turbine fouling and other related
problems illustrate the need to improve the understanding in this area.
Background
EPRI chemistry guidelines on all-volatile treatment (Report TR-105041), phosphate
treatment (TR-103665), oxygenated treatment (TR-102285) and caustic treatment (TR104007) provided direction and guidance on the use of these chemistries to control
drum boiler water and steam chemistry. A further report (TR-105040) provides
guidance on selection and optimization of feedwater and boiler water treatments. In
spite of the guidance provided by these documents, it was apparent that a better
understanding of copper corrosion, transport and deposition phenomena was needed.
Objective
To characterize the state-of-knowledge of copper alloys and copper oxides in fossil
plant cycles.
Approach
The authors developed a draft State-of-Knowledge document based on relevant
information published in the worldwide literature. The draft document was sent to
members of the EPRI Fossil Plant Cycle Chemistry Group for review and comment.
Comments received were considered by the authors and appropriate revisions were
made prior to publication of the final report.
Results
This state of knowledge document provides comprehensive information on all aspects
of copper alloys and the relevant copper oxides in fossil plant cycles:
Corrosion in high purity and contaminated feedwater
Electrochemistry in feedwater environments
Transport of copper and its oxides in feedwater, boiler water and steam
iii

Volatility and solubility of copper and its oxides in steam


Deposition phenomena in the steam turbine.
The document also provides a section on the known solutions to the various problems,
and a section on step-by-step procedures for assessing and evaluating the problems and
implementing corrective actions. The final section outlines the deficiencies in the state
of knowledge and where further R&D is required.
EPRI Perspective
This is the first time the underlying science for all aspects of copper in the fossil plant
cycle has been assembled. The section on electrochemistry will be further
comprehensively reviewed in a future document.
One of the main conclusions from this work is that the lack of understanding in each of
the areas prevents EPRI from currently developing a guideline to ensure that turbine
copper deposition problems will not occur in a specific unit. However, there is
sufficient experience around the world to delineate a list of Key Contributing Factors
to this problem. Also a set of Guiding Principles has been developed which if
followed by the utility industry will minimize and may eliminate the problem.
Further work in each of the areas is planned within the Boiler and Turbine Steam and
Cycle Chemistry Target.
TR-108460
Interest Categories
Fossil steam plant performance optimization
Fossil steam plant O&M cost reduction
Applied science and technology
Keywords
Power plant availability
Power plant performance
Steam turbines
Deposition
Feedwater
Copper

iv

EPRI Licensed Material

ABSTRACT
Recently utility users have experienced numerous problems resulting from copper alloy
corrosion, and the transport of copper and its oxides around the fossil plant cycle. A
number of serious incidences of copper/copper oxide deposition in the HP turbine
have also occurred.
This State-of-Knowledge Report was prepared as the first step in Program Copper,
initiated by EPRI to improve the scientific and practical understanding of copper alloys
under fossil plant unit service conditions. Its purpose is threefold. The report
summarizes the current understanding of copper corrosion, transport and deposition
activity. Also, it identifies deficiencies in the existing knowledge base and suggests
research and development activity needed to improve the situation. Finally, and most
importantly, it outlines approaches and strategies for use by utility personnel in
defining, evaluating and responding to ongoing and potential future copper transport
problems in fossil plant units.
Despite deficiencies in the understanding that have been identified, there is sufficient
information to provide guidance for utilities on developing cycle chemistries to
minimize the problems that can exist with copper in the cycle. Based on the experience
base, a set of Guiding Principles has been developed.
Subsequent research efforts to be conducted as part of Program Copper will improve
the understanding of the science of copper corrosion, electrochemistry, volatility,
solubility and transport. Application of the information resulting from these activities
will lead to refinement of cycle chemistry guideline limits, the Guiding Principles, and
the actions to be followed by utility personnel.

EPRI Licensed Material

ACKNOWLEDGMENTS
The authors acknowledge the review and contributions from members of the EPRI
Fossil Plant Cycle Chemistry Group. The final draft of the guidelines was reviewed by members
of this group.
Extensive comments and suggestions were received from the following:
Mr. A. Aschoff
Mr. M. Ball
Dr. J. Bellows
Mr. D. E. Hubbard
Mr. O. Jonas
Mr. A. Lindberg
Mr. J. Matthews
Dr. A. Petrov
Mr. F. Pocock

EPRI Consultant
EPRI Consultant, UK
Westinghouse
American Electric Power
Jonas Inc.
ComEd.
Duke Power
Moscow Power Institute
EPRI Consultant

The authors also wish to express their appreciation and gratitude to members of the
clerical staff at Sheppard T. Powell Associates, LLC who were involved in word
processing for the early draft, and to Mr. Daniel G. Whittemore for preparing many of
the original graphics. Lorrain Sargent and Jeff Sargent, of Pacific Publications, finalized
the word processing and document preparation.

vii

EPRI Licensed Material

CONTENTS

1 INTRODUCTION .......................................................................................................1-1
OVERVIEW OF THE EPRI FOSSIL PLANT CYCLE CHEMISTRY PROGRAM.......1-1

Volatility of Salts in Steam Cycles..........................................................................1-2


Phosphate Chemistry/Hideout/Corrosion ..............................................................1-3
Deposition and Chemical Cleaning ........................................................................1-5
Steam, Chemistry and Corrosion in the Phase Transition Zone (PTZ)..................1-6
EPRI FOSSIL PLANT GUIDELINES AND MANAGEMENT APPROACHES FOR
CYCLE CHEMISTRY ................................................................................................1-6
FEEDWATER TREATMENT ....................................................................................1-7
All-Ferrous Feedwater Systems ............................................................................1-7
MIXED METALLURGY FEEDWATER SYSTEMS (COPPER CONTAINING) ..........1-9
Alternative Water Treatment Chemicals ..............................................................1-11
GUIDELINE DEVELOPMENTS LEAD TO SELECTION AND OPTIMIZATION ......1-12
PROGRAM COPPER .............................................................................................1-13
REFERENCES........................................................................................................1-15
2 INTRODUCTION TO COPPER ALLOYS AND UTILITY INDUSTRY EXPERIENCE2-1
COPPER AND COPPER ALLOYS ...........................................................................2-1
EARLY UTILITY EXPERIENCE WITH COPPER ALLOYS.......................................2-6
USE OF COPPER ALLOY HEATERS IN FOSSIL UNITS WITH DRUM BOILERS..2-7
PROBLEMS RESULTING FROM USE OF COPPER ALLOYS IN FOSSIL UNITS
WITH DRUM BOILERS ............................................................................................2-8

Deposition in High Pressure Heaters and on Boiler Waterwalls ..........................2-10


Deposition in Turbines .........................................................................................2-13
UTILITY EXPERIENCE AND CONCERNS WITH TURBINE COPPER FOULING.2-15
Identification of Contributing Factors for Turbine Copper Fouling .......................2-15
Guiding Principles ................................................................................................2-16
REFERENCES........................................................................................................2-18
ix

EPRI Licensed Material

3 CORROSION OF COPPER ALLOYS IN FEEDWATER HEATERS .........................3-1


INTRODUCTION ......................................................................................................3-1
COPPER OXIDE FORMATION AND STABILITY .....................................................3-4
FACTORS INFLUENCING COPPER CORROSION AND RELEASE ACTIVITY......3-8

Effect of Oxygen ..................................................................................................3-11


Effect of pH ..........................................................................................................3-13
Effect of Velocity ..................................................................................................3-17
Effect of Temperature ..........................................................................................3-20
CONCLUDING REMARKS ON THE CORROSION OF COPPER ALLOYS ...........3-21
REFERENCES........................................................................................................3-22
4 ELECTROCHEMISTRY OF COPPER ......................................................................4-1
INTRODUCTION ......................................................................................................4-1
POTENTIAL-pH DIAGRAMS FOR COPPER............................................................4-2

High Temperature Potential-pH Diagrams for Copper ...........................................4-7


Concluding Remarks on Potential-pH Diagrams ...................................................4-9
COPPER CORROSION KINETICS ........................................................................4-10
Anodic Partial Reactions......................................................................................4-13
Cathodic Partial Reactions ..................................................................................4-16
PASSIVITY OF COPPER .......................................................................................4-19
Field Observations...............................................................................................4-20
Mechanisms of Passivation and Passivity Breakdown ........................................4-22
CORROSION MONITORING BY ELECTROCHEMICAL TECHNIQUES ...............4-31
ENEL Experience ................................................................................................4-32
CONCLUDING REMARKS ON THE ELECTROCHEMISTRY OF COPPER ALLOYS
IN FEEDWATER.....................................................................................................4-34
REFERENCES........................................................................................................4-34
5 COPPER TRANSPORT IN FEEDWATER SYSTEMS ..............................................5-1
INTRODUCTION ......................................................................................................5-1
SOURCES OF COPPER ..........................................................................................5-2

Copper Release from Low Pressure Heaters ........................................................5-2


Copper Release from High Pressure Heaters .......................................................5-4
FORM OF COPPER .................................................................................................5-5
EFFECT OF PEAKING AND CYCLING SERVICE ON COPPER TRANSPORT......5-8
MINIMIZING COPPER TRANSPORT.....................................................................5-14
x

EPRI Licensed Material

MONITORING COPPER TRANSPORT ACTIVITY.................................................5-16


REFERENCES........................................................................................................5-18
6 VOLATILITY AND SOLUBILITY OF COPPER IN STEAM.......................................6-1
INTRODUCTION ......................................................................................................6-1
CARRYOVER IN DRUM BOILERS...........................................................................6-1

Mechanical Carryover............................................................................................6-1
Volatile Carryover ..................................................................................................6-3
Carryover Testing ..................................................................................................6-3
VOLATILITY OF COPPER IN DRUM BOILERS .......................................................6-4
Field Experience ....................................................................................................6-7
SOLUBILITY OF COPPER IN STEAM .....................................................................6-9
CONCLUDING REMARKS .....................................................................................6-10
REFERENCES........................................................................................................6-11
7 COPPER DEPOSITION IN TURBINES.....................................................................7-1
INTRODUCTION ......................................................................................................7-1
TRANSPORT TO THE STEAM PATH AND TURBINE.............................................7-3

Importance of Individual Transport Mechanisms ...................................................7-3


DEPOSITION IN TURBINES ....................................................................................7-6
Effect of Deposits on Capacity...............................................................................7-7
Turbine Copper Deposits and Steam Path Audits .................................................7-8
Deposit Characteristics ..........................................................................................7-9
Mechanism of Deposition ....................................................................................7-14
CONCLUDING REMARKS ON COPPER DEPOSITION IN TURBINES ................7-17
REFERENCES........................................................................................................7-18
8 SOLUTIONS FOR COPPER PROBLEMS ................................................................8-1
INTRODUCTION ......................................................................................................8-1
A. DEPOSIT REMOVAL (SHORT TERM SOLUTIONS)...........................................8-2

Solution 1. Cleaning Turbine .................................................................................8-4


B. CHEMISTRY APPROACHES ...............................................................................8-7
Solution 4. Optimize Feedwater Chemistry for Mixed-Metallurgy Systems ..........8-7
Step 1Review of Water Chemistry, Operation, and Experience ....................8-10
Step 2Baseline Monitoring .............................................................................8-11
Step 3Water Chemistry Optimization .............................................................8-12
xi

EPRI Licensed Material

Step 4Design and Material Changes .............................................................8-13


Step 5Operation.............................................................................................8-13
Step 6Monitoring to Compare with Baseline Data .........................................8-13
Step 7Normal Operation and Monitoring .......................................................8-14
Step 8Continual Check of Chemistry .............................................................8-14
Step 9Long-Term Plans .................................................................................8-14
Solutions 6 and 7. Develop layup/shutdown/startup procedures. .......................8-14
C. EQUIPMENT CHANGES...................................................................................8-15
Solution 8. Change Feedwater Heater Materials ................................................8-15
Feedwater Heater Replacement .........................................................................8-15
Feedwater Heater Tube Material Comparison ....................................................8-18
Examples of Replacing Tube Material in Feedwater Heaters ..............................8-22
Solutions 9-11. Modify Unit and Operating Procedures ......................................8-28
REFERENCES........................................................................................................8-29
9 ASSESSMENT AND CORRECTION OF COPPER CORROSION, TRANSPORT AND
DEPOSITION PROBLEMS ..........................................................................................9-1
INTRODUCTION ......................................................................................................9-1
OVERVIEW OF UTILITY ACTIONS .........................................................................9-1

Action 1 - Determine Unit Susceptibility to Copper Problems ................................9-2


Action 2 - Evaluate Condition of Copper Alloy Components..................................9-8
Action 3 - Characterize Copper in Cycle Deposits .................................................9-8
Action 4 - Optimize Cycle Chemistry ...................................................................9-11
Action 5 - Improve Startup Procedures and Layup Practices ..............................9-12
Action 6 - Examine Cycle Design.........................................................................9-15
Action 7 - Follow-up Monitoring Campaigns ........................................................9-16
SOME CONCLUDING REMARKS ON CONDUCTING MONITORING
CAMPAIGNS FOR ASSESSMENT OF COPPER TRANSPORT............................9-16
Sample Points .....................................................................................................9-18
Parameters to be Monitored ................................................................................9-18
Sampling and Analysis of Corrosion Products .....................................................9-18
Guidelines for Evaluation of Copper Transport Data ...........................................9-19
RESOURCES FOR COPPER IN UTILITY PLANT CYCLES ..................................9-20
REFERENCES........................................................................................................9-21

xii

EPRI Licensed Material

10 FUTURE RESEARCH REQUIREMENTS .............................................................10-1


REFERENCES........................................................................................................10-3

xiii

EPRI Licensed Material

LIST OF FIGURES
Figure 1-1 Overview of the development of the EPRI Fossil Plant Cycle Chemistry
Guidelines ......................................................................................................................1-2
Figure 1-2 Partitioning Constants KD for Common Boiler Water Salts, Acids and
Bases Represented by Mathematical Functions of the Reciprocal of
Temperature in Kelvin up to the Critical Temperature of Water, Tc........................1-4
Figure 2-1 Phase diagram for the Cu-Zn system. .............................................................2-4
Figure 2-2 Phase diagram for the Cu-Al system. ..............................................................2-5
Figure 2-3 Phase diagram for the Cu-Ni system...............................................................2-6
Figure 2-4 Typical Locations of Impurity Ingress, Corrosion and Deposition in
a Drum Cycle...............................................................................................................2-10
Figure 25A Multilayered Deposit with Metallic Copper in Pits..................................2-12
Figure 25B Multilayered Deposit with Metallic Copper Plated on Metal Surface ....2-12
Figure 26 Effect of Internal Scale Thickness on Tube Metal Temperature .................2-13
Figure 27 Turbine Copper Deposits ...............................................................................2-14
Figure 3-1 Average Copper Release as a Function of Feedwater pH for Various
Copper Alloys................................................................................................................3-2
Figure 3-2 Corrosion Product Release in ppb for Carbon Steels as a Function of
Feedwater pH ................................................................................................................3-3
Figure 3-3 Schematic Cross-sectional Diagram of the Structure and Charge Mobility
within a Copper Oxide Layer Formed on Pure Copper Metal in Aqueous
Solution ..........................................................................................................................3-7
xv

EPRI Licensed Material

Figure 3-4 Average Copper Corrosion Product Transport Rates for Normal
Operation at Fox Lake Unit 3 .....................................................................................3-10
Figure 3-5 Corrosion Products at the Economizer Inlet and in the Reheat Steam
for Typical Operation at Newton Unit 1 ...................................................................3-10
Figure 3-6 Steady State Metal Release Rates in Neutral Water Flowing at 5 ft/s
Containing (a) Low Oxygen (3 g/kg O2 with 100200 g/kg H2) and (b)
High Oxygen (200 g/kg O2 with 25 g/kg H2)......................................................3-12
Figure 3-7 Effect of Hydrazine Concentration on Feedwater Copper Transport
at pH 8.8 in an Australian Unit..................................................................................3-13
Figure 3-8 Effect of pH on Copper Release Rate From 71Cu/28Zn/Sn (SoMs-71
brass, or Admiralty brass) Exposed to Water Containing Carbon Dioxide
(for pH <7) or ammonia (for pH>7) at Temperatures of 30C and 90C. ..............3-14
Figure 3-9 Effect of pH on Steady State Release Rates for 90Cu/10Ni and
70Cu/30Ni Exposed to Ammonia Solutions Containing 8 - 21 mg/kg
Oxygen, Flowing at ~ 1 ft/s (0.3 m/s) and at a Temperature of 35 - 38C. ........3-15
Figure 3-10 Corrosion Rates for Copper Alloys Exposed to Ammonia Solutions
Containing 100 200 mg/kg Oxygen* at 40C. pH 7, 9.4 and 10 Solutions
were Obtained by Adding 0, 2, and 20 mg/kg NH3 Respectively.........................3-16
Figure 3-11 Effects of Velocity and Deaeration on the Reduction of the Wall
Thickness of Admiralty Brass Tubes at 185C. (Water was aerated for the
first 2200 hours of exposure and deaerated thereafter.) ..........................................3-19
Figure 3-12 Effect of Nickel Content (0-100%) and Temperature (200 to 350C) on
the Corrosion Rate of Copper-Nickel Alloys Exposed for 250 Hours to
Oxygen-Bearing Water Under Saturated Pressure Conditions. .............................3-21
Figure 4-1 Potential-pH Diagram for the Copper-Water System at 25C.......................4-3
Figure 4-2 Potential-pH Diagram for the Copper-Water-Ammonia System at 25C ....4-7
Figure 4-3 Potential-pH diagram for Cu-H2O at 100C. Activity of dissolved copper
species = 10-6 mol/kg; pressure = 1 bar.......................................................................4-8
Figure 4-4 Potential-pH diagram for Cu-H2O at 250C. Activity of dissolved copper
species = 10-6 mol/kg; pressure = 1 bar.......................................................................4-9

xvi

EPRI Licensed Material

Figure 4-5 Current/Voltage Curves for Electrochemical Processes Rate-Limited


by (a) Charge Transfer and (b) Mass Transfer .........................................................4-11
Figure 4-6 Schematic Current/Voltage Curve for a Passive Metal ...............................4-20
Figure 4-7 Schematic of processes that lead to the formation of bilayer passive
films on metal surfaces. ..............................................................................................4-23
Figure 4-8 Schematic of processes that lead to the formation of the Cu2O layer and the
Cu(OH)2/CuO outer layer in the passive film on copper in oxidizing aqueous

'

media. VCu
= cuprous ion vacancy in Cu2O, CuCu = cuprous cation in Cu2O, VO
=
oxygen vacancy in Cu2O. This scheme assumes that the barrier layer is a cation
vacancy conductor; a comparable scheme can be constructed for an interstitialconducting barrier layer.. ...........................................................................................4-23

Figure 4-9 Cartoon outlining various stages of pit nucleation according to


the PDM. . ...................................................................................................................4-26
Figure 4-10 Summary of proposed reactions to passivity breakdown. ........................4-27
Figure 4-11 Effect of Chromium Additions on Seawater Impingement-Corrosion
Resistance of Copper-Nickel Alloys. 36-Day Test with 7.5 m/s Jet Velocity;
Seawater Temperature: 27C......................................................................................4-30
Figure 5-1 Copper Concentrations at Port Everglades Unit 4 During July 1985............5-6
Figure 5-2 Levels of Suspended and Dissolved Copper and Iron Present Under
Oxygenated Treatment .................................................................................................5-7
Figure 5-3 Feedwater Metals Concentrations in Feedwater During Startup
Operations......................................................................................................................5-9
Figure 5-4 Relationship Between Power and Heater Shell Pressure...............................5-9
Figure 5-5 Corrosion Product Sampler Devices used at Laramie River Station ..........5-11
Figure 5-6 Copper Concentrations During Normal Operation of Unit No. 2
at Laramie River Station .............................................................................................5-12
Figure 5-7 Corrosion Product Transport During Return to Service of Unit
No. 1 at Laramie River Station (April 1522, 1991)..................................................5-14
Figure 5-8 Copper, Nickel and Zinc Generation, Transport and Effects in a
Drum Boiler Unit with Mixed Metallurgy................................................................5-17
xvii

EPRI Licensed Material

Figure 6-1 Representative Drum Boiler Mechanical Carryover(3)....................................6-2


Figure 6-2 Distribution Ratios for Common Boiler Water Contaminants ......................6-4
Figure 6-3 Boiler Water Copper Concentration/Drum Boiler Pressure Control
Curves Developed at Miami Fort Station(21)................................................................6-8
Figure 6-4 Solubility of cuprous (Cu2O) and cupric (CuO) oxides in superheated
steam as a function of pressure..................................................................................6-11
Figure 7-1 Solubility of Cupric Oxide at Temperatures and Pressures
Corresponding to Various Stages of the High Pressure Turbine..............................7-2
Figure 7-2 Distribution of Copper in a Fossil Unit ...........................................................7-6
Figure 7-3 The effect of the elimination of N2H4 from the feedwater of a mixedmetallurgy feedwater system on the copper transport in feedwater and
saturated steam..............................................................................................................7-7
Figure 7-4 Decrease in gross MW capacity after eliminating the oxygen scavenger
N2H4 from the feedwater. Note that the MW keeps decreasing even after
the N2H4 has been restored (A)....................................................................................7-8
Figure 7-5 Effect of Nozzle Area Change on Flow Capacity of Impulse-Type
Turbines(9). (This figure is also applicable to the first stage of reaction-type
turbines) .........................................................................................................................7-9
Figure 7-6 Solubility of Cupric Oxide in Steam as a Function of Specific Volume(3) ...7-15
Figure 8-1 Road Map for Optimizing Feedwater Treatment for Mixed Metal
Systems (Cu/Fe)............................................................................................................8-9
Figure 9-1 Summary Road Map for Utility Assessment and Correction of Problems
Involving Copper Transport ........................................................................................9-3
Figure 9-2 Road Map for Determination of Unit Susceptibility to Copper Problems:
Action 1A, Review of Experience ................................................................................9-6
Figure 9-3 Road Map for Determination of Unit Susceptibility to Copper
Problems - Action 1B, Review of Design and Operating Characteristics ................9-7
Figure 9-4 Road Map for Action 2, Evaluate Condition of Copper Alloy
Components...................................................................................................................9-9
Figure 9-5 Road Map for Action 3, Characterize Copper in Cycle Deposits................9-10
xviii

EPRI Licensed Material

Figure 9-6 Road Map for Action 4, Optimize Cycle Chemistry.....................................9-12


Figure 9-7 Road Map for Action 5, Review Startup Procedures and
Layup Practices ...........................................................................................................9-14
Figure 9-8 Road Map for Action 6, Examine Cycle Design ...........................................9-17

xix

EPRI Licensed Material

LIST OF TABLES
Table 1-1 "Core" Monitoring Parameters (Minimum level of instruments for all
plants/units) ..................................................................................................................1-8
Table 1-2 Generation of Feedwater Corrosion Products by Corrosion and FlowAccelerated Corrosion, and the Major Unit Transport and Deposition Problem
Areas for All-Ferrous Systems .....................................................................................1-9
Table 1-3 Generation of Feedwater Corrosion Products by Corrosion and FlowAccelerated Corrosion and the Major Unit Transport and Deposition Problem
Areas for Mixed Metallurgy Systems........................................................................1-10
Table 2-1 Properties of Copper, Silver, and Gold ............................................................2-1
Table 2-2 Nominal Composition of Commercial Cu-Base Alloys Used in
Feedwater Heater Systems ...........................................................................................2-2
Table 2-3 Properties of Some Copper Base Alloys...........................................................2-3
Table 2-4 Copper Alloys Employed as Condenser and Feedwater Heater Tubing
in Fossil Plant Cycles ....................................................................................................2-8
Table 2-5 Heater Pressure Drop Values Prior to and Following Copper Removal ....2-11
Table 2-6 Normal Cycle Chemistry Limits at the Economizer Inlet for Units with
Mixed-Metallurgy Feedwater Systems .....................................................................2-17
Table 3-1 Corrosion Mechanisms Influencing Copper Alloys in Feedwater Heater
Service (a) Water Side, and (b) Steam Side .................................................................3-4
Table 3-2 Chemical, Structural, and Thermodynamic Properties of the Oxides
and Hydroxide of Copper ............................................................................................3-5
Table 3-3 Forms of Copper which may be Present in Feedwater...................................3-9
Table 3-4 The Influence of Addition of 400 ppm (NH4)2CO3 on the Corrosion
Rate of Selected Copper-Based Alloys in an Ammoniacal Solution
(500 ppm NH3) at Ambient Temperature .................................................................3-17
xxi

EPRI Licensed Material

Table 3-5 Heat Exchange Institute Criteria for Maximum Feedwater Heater Tube
Velocities ......................................................................................................................3-20
Table 4-1 Standard Electrode Potentials at 25C Versus Standard Hydrogen
Electrode ...................................................................................................................... 4-12
Table 4-2 Exchange Current Density Values ..................................................................4-13
Table 4-3 Available Techniques for Monitoring Corrosion and Potential...................4-32
Table 5-1 Effect of High Pressure Heater Drains on System Copper Levels...............5-10
Table 5-2 Mitigation of Copper Transport......................................................................5-15
Table 6-1 Solubility of Copper in Water and Distribution Ratio....................................6-6
Table 6-2 Influence of Ammonia on Distribution Ratios for Copper Oxide .................6-7
Table 6-3 Maximum Copper Solubilization at 4500 psi, 1150F .....................................6-9
Table 7-1 Comparison of Available Copper Transport Paths.........................................7-4
Table 7-2 Copper Species Identified by a Turbine Deposit Survey .............................7-10
Table 7-3 Forms of Copper Reported in Turbine Deposits from Drum Boiler
Units .............................................................................................................................7-10
Table 7-4 Turbine Deposit Composition: Northport Station, Unit No. 4.....................7-11
Table 7-5 Turbine Deposit Composition: Columbia Generating Station,
Units Nos. 1 and 2 .......................................................................................................7-12
Table 7-6 Turbine Deposit Composition: Miami Fort Station, Units Nos. 7 and 8 ....7-13
Table 7-7 Turbine Deposit Composition: Craig Station, Units Nos. 1 and 2...............7-14
Table 7-8 Isoelectric Point of Surface (IEPS) Values for Several Compounds.............7-17
Table 8-1 Possible Solutions for Copper Problems ..........................................................8-2
Table 8-2 Typical Concentration and Temperature of Use for Common Solvents
Used for Waterwall and Economizer Cleanings ........................................................8-3
Table 8-3 Capacity of Selected Solvents for Constituents in Waterwall and
Economizer Deposits ....................................................................................................8-3
xxii

EPRI Licensed Material

Table 8-4 Design Minimum Tube Wall Thickness.........................................................8-19


Table 8-5 Feedwater Heater Tubing and Properties......................................................8-20
Table 8-6 LP Feedwater Heater Comparison Chart.......................................................8-23
Table 8-7 HP Feedwater Heater Comparison Chart......................................................8-24
Table 8-8 Key Factors Involved in Replacing Admiralty Tubes with Stainless
Steel in a LP Heater (Data extracted from Tables 8-6 and 8-7) ...............................8-26
Table 8-9 Key Factors Involved in Replacing 70-30 Cu-Ni Tubes with Stainless
Steel in a HP Heater. (Data extracted from Tables 8-6 and 8-7) ............................8-27
Table 9-1 Goals for Copper in Mixed Metallurgy Drum Boiler Units .........................9-20

xxiii

EPRI Licensed Material

EXECUTIVE SUMMARY
The Interim Consensus Guidelines (ICG) for Fossil Plant Cycle Chemistry (CS-4629)
were introduced in 1986 to provide the guidance needed to better control cycle
corrosion and deposition. Introduction of comparable, unified cycle chemistry
guidelines by the then Central Electricity Generating Board (CEGB) in the United
Kingdom, the Vereinigung der Grosskraftwerks Betreiber (VGB) in Germany, and the
Central Research Institute of Electric Power Industry (CRIEPI) in Japan had all resulted
in significant reductions in cycle corrosion-related failures and improvements in unit
availability.
These ICG were primarily based on current industry experience and well established
thermodynamic data that pertain to water and steam chemistry. Sources used for this
information included equipment manufacturers, technical literature, and international
guideline specifications and standards. Additional input, assistance, and guidance was
provided by EPRI's Fossil Plant Cycle Chemistry Group and by visits to operating units
of 12 utilities. In addition, a selected group of 25 members of the EEI Chemistry
Committee reviewed a draft of the Guidelines.
The ICG were considered interim; they would be reviewed over the next several years
in the light of subsequent research and operating experience in implementing these
Guidelines, and then revised as necessary. The Guidelines were developed to take
differences in unit design into account to the greatest possible degree. Modification of
the Guidelines by individual utilities to reflect actual, plant specific characteristics and
local operating experiences was an inherent feature of their application.
The ICG were based upon the cycle chemistry control practices most common in the
U.S. in the early 1980s, specifically congruent phosphate treatment (CPT) or all-volatile
treatment (AVT) of boiler water with the deoxygenated condensate/feedwater being
treated with ammonium hydroxide and hydrazine. In the late 1980s and early 1990s
most utilities in the U.S. (over 80%) based their cycle chemistry on the ICG with
suitable modifications for individual units. A Guideline on Oxygenated Treatment
(OT) was introduced in the U.S. (TR-102285) and a new Guideline on Phosphate
Treatment (TR-103665) was the second addition/change to the original ICG. A new
Guideline on All-Volatile Treatment (TR-105041) represents the latest revision to the
ICG. A report describing international experience with caustic treatment of drum
boiler units has also been published (TR-104007). Just recently a new guidance

xxv

EPRI Licensed Material

document was published which outlines the steps necessary to determine the optimum
boiler water and feedwater treatments for units (TR-105040).
Much research has been conducted since the issuance of the ICG. A detailed
monitoring of four fossil plants in the U.S. and the collection of data from 31
international units led to the identification of a number of deficiencies in the basic
understanding of cycle chemistry:

Volatility of salts in steam

Phosphate treatment, hideout and corrosion

Deposition around the cycle

Steam chemistry and corrosion in the phase transition zone of the steam turbine

Each of these areas has been, or is currently being, addressed by EPRI.


While most utilities operating with AVT chemistry could meet the ICG limits, a number
of problem areas were identified which relate to the generation, transport and
deposition of feedwater corrosion products:

Flow-accelerated corrosion (FAC) of components in the feedwater and at the


economizer inlet

Fouling of components in the feedwater and boiler circuits

Deposition on boiler waterwalls and the need to chemical clean frequently

Boiler tube failures related to internal tube deposits

Deposition of copper and copper oxides on the HP turbine blades.

These relate to the specific chemistry-equipment design and metallurgy of the


feedwater system. With all-ferrous metallurgy the use of deoxygenated feedwater with
less than one ppb of oxygen at the economizer inlet and an excess of oxygen scavenger
(usually hydrazine) results in a very reducing environment (ORP 0 mV), which can
actually increase the amount of feedwater corrosion products. With a mixed feedwater
system there is always a balance needed for the ferrous and copper base alloys and this
must be developed by monitoring and testing, however it is clear that a reducing
environment is needed for copper based alloys to minimize copper corrosion and
transport.
The introduction of the oxygenated treatment in the U.S. has led to a much better
understanding of the feedwater chemistry-equipment environment in all-ferrous
metallurgy units. Some of the ancillary results have also provided guidance on the use
of oxygen scavengers and the development of the optimum feedwater treatment.
Application of EPRI chemistry guidelines in drum units with mixed metallurgy
feedwater systems has been beneficial yet performance losses related to corrosion,
xxvi

EPRI Licensed Material

transport and deposition of copper continue to be experienced in many fossil units.


Although virtually any component between the last low pressure heater and the HP
turbine may be subject to copper deposition, it is fouling of the high pressure section of
the turbine which is most costly. In order to enable utilities to effectively identify
opportunities to minimize copper transport and associated problems, EPRI initiated
Program Copper, a series of projects intended to assess the existing state-ofknowledge and conduct research targeted to improve the understanding in areas
considered deficient.
This State-of-Knowledge Report is the first document published under Program
Copper. Following a review of problems encountered in fossil plant units, summary
discussions of various aspects of the science of copper in fossil plant cycles are
presented, based on published information which reflects worldwide experience on
corrosion, electrochemistry, transport, solubility, volatility, deposition, and the possible
solutions. Using the current state-of-knowledge, a series of road maps is provided for
use by utilities needing to evaluate copper transport potential in existing units or to
minimize it in cases where performance penalties have already been incurred. This
information is supplementary to the current EPRI guidelines on boiler water chemistry
and feedwater treatment, and is complementary to the Selection and Optimization of
Boilerwater and Feedwater document.
PURPOSE AND SCOPE
The purposes of this State-of-Knowledge Report are a) to provide a single reference
source which describes various scientific aspects of copper transport activity as may be
experienced in fossil plant units with mixed metallurgy feedwater systems, b) to
provide guidance to utilities with mixed metallurgy feedwater systems on definition,
assessment and reduction of copper transport activity and, c) to identify deficiencies in
the current state-of-knowledge which require future research.
The State-of-Knowledge Report is organized and presented in a manner which should
facilitate assimilation of important aspects of the science of copper transport activity by
plant personnel and provide clear direction with respect to investigative, assessment
and corrective actions which are available for site-specific consideration and
implementation. The principles and concepts presented may be applied to any drum
boiler unit with copper alloy feedwater heater and/or condenser tubes.
COPPER TRANSPORT ACTIVITY ASSESSMENTS
The process of evaluating copper transport activity in specific fossil units is presented
as a series of seven actions in Section 9.

Action 1 - Determine Unit Susceptibility to Copper Problems

Action 2 - Evaluate Condition of Copper Alloy Components


xxvii

EPRI Licensed Material

Action 3 - Characterize Copper in Cycle Deposits

Action 4 - Optimize Cycle Chemistry - Conduct Monitoring Campaign

Action 5 - Improve Startup and Layup Procedures

Action 6 - Examine Cycle Design

Action 7 - Final Monitoring Campaign

Text discussion of the actions is supplemented by a series of road maps. Depending on


operating experience and available data, it may be desirable to implement actions in
series, in parallel or in a different order than suggested in the report.
Many related aspects of cycle chemistry are not addressed in detail since they are
already covered in detail in existing EPRI publications. However, reasonable effort has
been made to include essential information; also, appropriate references and crossreferences are indicated.
Section 8 delineates the main solutions to the various problems and features of having
copper alloys in the cycle. Documentation of costs related to copper transport activity
and assessment of available improvements and benefits to be derived as a result of
following the suggested actions remain the responsibility of utility personnel, since
conclusions derived as the result of such evaluations are very specific to the unit under
consideration. However, the report does include relevant published experience of
utilities that have implemented changes to reduce copper transport.
OUTLINE
The State-of-Knowledge Report consists of ten major sections. Section 1 provides an
overview of the EPRI Cycle Chemistry Program and introduces Program Copper.
Section 2 introduces copper alloys and identifies the nature of problems which may be
encountered in units with mixed metallurgy feedwater systems as a consequence of
excessive corrosion product transport activity. Utility and EPRI responses to turbine
copper fouling and related problems are also presented. Section 2 also provides the
current EPRI guideline limits for feedwater in mixed-metallurgy systems, and a set of
Guiding Principles, which, if followed, will lead to successful operation of units with
copper alloys. The science of copper corrosion, with emphasis on feedwater heater
conditions, is the subject of Section 3. Electrochemical principles, including
thermodynamic and kinetic aspects, as available and applicable to copper systems are
discussed in Section 4. Section 5 provides a summary of worldwide experience with
copper transport in fossil units over the last five decades. Volatility and solubility
properties, which influence copper in the steam path, are the subject of Section 6.
Deposition of copper in turbines is covered in Section 7. Section 8 brings together the
available solutions for copper problems in the cycle in three major categories: deposit
removal, chemistry approaches, and equipment changes. Section 9 outlines utility
actions available to reduce copper transport, and identifies other relevant resources
xxviii

EPRI Licensed Material

available from EPRI. Section 10 reviews data needs and other deficiencies in the
existing state-of-knowledge which would benefit from further research and
development activity.

xxix

EPRI Licensed Material

1
INTRODUCTION

OVERVIEW OF THE EPRI FOSSIL PLANT CYCLE CHEMISTRY PROGRAM


The Electric Power Research Institute (EPRI) Fossil Plant Cycle Chemistry Program has
the following goals:

To eliminate boiler tube failures related to cycle chemistry

To eliminate turbine chemical problems (low-pressure blade and disk cracks, and
serious deposits throughout the turbine)

To eliminate the need for boiler chemical cleaning

To identify simple, reliable cycle chemistry instrumentation:

minimum levels of instrumentation needed for each unit, and the associated
quality assurance (QA) / quality control (QC) procedures, and
direct on-line, in-situ instruments

To shorten the startup period by:

optimization of shutdown, lay-up and startup chemistry

the elimination of unnecessary chemical holds in the startup sequence

To develop operational guidelines with action levels for all units, and

To identify the optimum managerial approach in support of cycle chemistry

The program was initiated with the development of the Interim Consensus Guidelines
(ICG) in 1986(1). An overview of the subsequent cycle chemistry guideline
developments is shown in Figure 1-1. The next additions to this will be the
Cycling/Startup/Shutdown/Layup Guidelines and Project Copper.
The ICG was followed by a detailed monitoring program at four U.S. fossil plants (2, 3).
Information was also collected at many international plants(4). The information that
was developed from these efforts led to the identification of four areas where the
understanding was deficient in the ICG:

Volatility of salts and how impurities partition between boiler water and steam
Phosphate chemistry/hideout/corrosion
Deposition around the cycle
Steam chemistry and corrosion in the phase transition zone (PTZ)
1-1

EPRI Licensed Material


Introduction

EPRI projects have been initiated in response to these areas of deficiency. Brief
information on these studies is included below together with information from other
studies.

Figure 1-1 Overview of the development of the EPRI Fossil Plant Cycle Chemistry
Guidelines

Volatility of Salts in Steam Cycles


Historically the ray diagram has provided a rough estimate for determining
vaporous carryover from the boiler water. But it was confirmed from plant
monitoring(2, 3) that chloride and sulfate concentrations can be as much as two orders of
magnitude higher in the steam than shown in the ray diagram.
To develop a more thorough understanding of the volatility of salts in steam cycles, the
EPRI research in this area began by investigating the partitioning of ammonium
chloride (NH4Cl) in laboratory-scale experiments from 120C (248F) to 350C (662F)(5).
This research revealed that while the dominant chloride species for NH4Cl solutions in
both high and low temperature liquid were NH4+ and Cl-, the species transported to
the equilibrated vapor were predominantly HCl and NH3. An approximately similar
1-2

EPRI Licensed Material


Introduction

picture has emerged from the continued research on the partitioning from sulfate
solutions in the presence of sodium and ammonium cations, although the hydrolysis
reactions of sulfate ion complicate the speciation. The significant species in solution at
low temperature (condensate, blowdown) under AVT conditions are ammonium ions,
ammonia, and hydroxide ions, with impurities of sulfur being present in the form of
sulfate ions. At boiler operating conditions, equilibrium thermodynamics dictate that
ammonia predominates over ammonium ion, whereas bisulfate and sulfate ions are at
much lower, but similar, concentrations. In the high temperature steam phase at
equilibrium with this solution, again ammonia predominates over HCl. At lower, but
comparable, concentration levels are ammonium chloride, sodium hydroxide, sulfuric
acid, sodium bisulfate and ammonium bisulfate, depending on the relative levels of
these impurities in the boiler water. These preliminary calculations predict further
rearrangement of the relative concentrations of the predominate molecules as the steam
cools, with a much larger range in values. Clearly the situation is more complex with
the addition of more potentially-volatile species, particularly those which undergo
additional reactions in the liquid phase, and this complex chemistry goes far beyond
that which can be predicted from the ray diagram. The partitioning constants for the
salts, acids and bases mentioned here are shown in Figure 1-2, where the partitioning
constant, KD, can be defined for a simple 1:1 electrolyte as the ratio of the concentration
of the neutral molecule in the vapor phase to the activities of the component ions in the
liquid phase.

Phosphate Chemistry/Hideout/Corrosion
Up to the mid 1990s many utilities continued to experience phosphate hideout and
boiler water pH instability when following the information on congruent phosphate
treatment (CPT) in the ICG(1). Also, an increasing number of utilities experienced
serious internal corrosion (attributed to acid phosphate corrosion) of the boiler
waterwalls and subsequent boiler tube failures when using this phosphate chemistry(6,
7). The sodium iron phosphate compound, maricite, has been found to be a magnetitephosphate reaction product associated with cases of serious corrosion, and a
distinguishing difference from caustic gouging.

1-3

EPRI Licensed Material


Introduction

Figure 1-2 Partitioning Constants KD for Common Boiler Water Salts, Acids and Bases
Represented by Mathematical Functions of the Reciprocal of Temperature in Kelvin up
to the Critical Temperature of Water, Tc.
An EPRI project(8, 9) was initiated to answer questions related to boiler tube corrosion
and phosphate hide-out that have occurred in some boilers operating under CPT and
to assist in modifying the ICG. This work generally extended the results of the 1964-68
ASME Test Program(10). The results are in general agreement with the literature
published on this subject. Specifically, no evidence of major corrosion attack was found
using phosphate based boiler water treatment under conditions of:

Saturation pressure of 2800 psig and heat flux up to 200,000 BTU/ft2

Departure from nucleate boiling (DNB) 12 hours in duration

Phosphate concentrations to 10 ppm

Sodium to phosphate molar ratios ranging from 1.84.0

Magnetite deposition of 4 mg/cm2 (~4 grams/ft2)

Low chloride and silica contamination

1-4

EPRI Licensed Material


Introduction

The results provide support for treatment methods which permit low levels (generally
<1 ppm) of free caustic, such as equilibrium phosphate treatment under the tube
cleanliness conditions tested.
Work conducted by the Canadian Electrical Association(11) identified the sodium-iron
phosphate reactions that take place up to 360C. The major iron reaction products that
cause hideout (or more specifically in these experiments, uptake by magnetite at
Na/PO4 molar ratios near 2.5) were identified from batch experiments as NaFe++PO4
(maricite) and Na4Fe+++(OH)(PO4)2 1/3NaOH. At higher Na/PO4 ratios
Na3-2xFex++PO4 (a solid solution with Na3PO4) replaces maricite as the stable reaction
product. At 360C (680F) uptake by magnetite behavior is similar except that there
appears to be no significant amount of iron (+2) reaction products with Na/PO4 ratios
of 2.5 or greater. If the Na/PO4 ratio is large (>3.5), no uptake by magnetite takes
place.
Nickel (NiO) reportedly behaves similarly. The Na/PO4 ratio in boiler water required
to avoid the formation of more acidic phosphate mixtures (maricite + iron III phases)
increases from about 2.3 at 315C (599F) to about 2.7 at 360C (680F). The injection of
solutions with Na/PO4 ratios above 3.0 causes little or no iron-containing phosphate
deposit to form at 360C (680F).
EPRI published a revised Guideline(12) for phosphate treatment for drum units which
took into account the results of all these studies and relevant utility experiences. This
has been accomplished by providing two phosphate treatments: the first called
phosphate treatment (PT) involves a broadening of the control range above the sodiumto-phosphate 2.8 molar ratio curve and allows operation with up to 1 ppm of free
sodium hydroxide; the second, equilibrium phosphate treatment (EPT) operates at or
below phosphate levels which would lead to hideout. In high performance units with
low tolerance for phosphate, operation with up to 1 ppm of free hydroxide is allowed.
The major philosophy change incorporated has been to try to minimize or eliminate
phosphate hideout and the continual correction of the boiler chemistry by addition of
the phosphate treatments (di, and mono-sodium phosphate). PT is essentially an
extension of EPT at higher phosphate levels. From a control viewpoint, the major
difference is in the level of allowed contaminants, which must be consistent with the
buffering capacity of the treatment in use. Since its introduction in 1994, the incidence
of corrosion has decreased markedly and utilities are able to control the phosphate
chemistry with minimum or reduced levels of hideout.

Deposition and Chemical Cleaning


Deposition has a very important influence on waterside failure mechanisms and
component performance. The deposition of feedwater corrosion products, and
particularly their minimization, on the waterwalls of the boiler is key to a successful
boiler treatment program. EPRI has recently initiated an exploratory project to develop
1-5

EPRI Licensed Material


Introduction

quantitative understanding of deposition processes throughout the steam and water


cycle. In the interim there are a number of published documents, which relate to the
operation of an optimum cycle chemistry program(1215), to the minimization of
deposition(16, 17) and to the determination of the need to chemically clean a boiler(18).

Steam, Chemistry and Corrosion in the Phase Transition Zone (PTZ)


Recently EPRI published a State-of-Knowledge document in this area(16) which included
information on steam chemistry, moisture nucleation, early condensate and deposition
within the phase transition area of the steam turbine. This work led to the formation of
an international collaboration consisting of 23 organizations that are performing
detailed monitoring of these areas in operating turbines. It is anticipated that the work
will lead to a complete understanding of the environment in the PTZ, which will
ultimately provide better steam chemistry limits.
EPRI FOSSIL PLANT GUIDELINES AND MANAGEMENT APPROACHES FOR
CYCLE CHEMISTRY
Over the period 19931996, EPRI has incorporated the information from all the ongoing cycle chemistry projects into individual guidelines which are revisions of the
Interim Consensus Guidelines:

Phosphate treatment guidelines to cover phosphate treatment (PT) and equilibrium


phosphate treatment (EPT)(12)

Oxygenated treatment for once-through and drum units(14)

All-volatile treatment for once-through and drum units(15)

EPRI has also prepared a document(19) which summarizes the worldwide experience
with caustic treatment for drum boilers. This treatment is currently utilized
successfully in over 50,000 MW of drum boilers at applied concentrations up to 2 ppm
NaOH.
The Selection and Optimization of Boiler Water and Feedwater was published in
1997(22). This document is the glue which brings the four guidelines together; it helps
a utility to select the optimum treatment for specific units and provides a road-map
methodology to optimize the feedwater for all-ferrous and mixed metallurgy feedwater
systems.
In parallel to the guidelines development, EPRI has developed a very successful Cycle
Chemistry Improvement Program(13, 20) and has demonstrated it with nine utilities. This
includes the minimum level of instrumentation that all fossil plants are considered to
need (core parameters) and which was developed as a result of the monitoring
program(2) and international data(4). Table 1-1 shows these parameters/instrumentation
together with the diagnostic parameters, which should be used (a) in cases of
1-6

EPRI Licensed Material


Introduction

contaminant ingress or when target levels are exceeded (troubleshooting parameters),


and (b) during commissioning of cycle chemistry.
Also included in the CCIP is the optimum management for a utilitys cycle chemistry
program and the methodology to record the costs/benefits of an improving cycle
chemistry.
FEEDWATER TREATMENT

All-Ferrous Feedwater Systems


Prior to the introduction of OT into the U.S., and in an attempt to reduce feedwater
corrosion products, operating guidelines, including EPRIs Interim Consensus
Guidelines(1), indicated that the feedwater should be deoxygenated (or reducing with
ORP<0 mV) and that oxygen levels should be less than 5 ppb. Over the last 10 years
some utilities have been striving to reduce air in-leakage but have not found that the
key indicators (such as time between chemical cleans) have markedly changed(21); this
indicates that there has not been a serious reduction in feedwater corrosion product
transport. In the same time period, utilities have been adding, increasing, and
changing oxygen scavengers. The result has generally been a reduction in economizer
inlet oxygen levels and a concomitant oxygen scavenger increase. Together, these
result in severe reducing conditions (oxidizing-reducing potentials (ORP) of less than 350 mV) in the feedwater leading to an increase in feedwater corrosion products in allferrous systems: the opposite result to the initial aspiration(17). Table 1-2 provides an
indication of the typical areas suffering from corrosion, flow-accelerated corrosion
(FAC) and generating feedwater corrosion products, as well as the major cycle problem
areas.

1-7

EPRI Licensed Material


Introduction

Table 1-1
Core Monitoring Parameters (Minimum level of instruments for all plants/units)
Parameters

Measurement Locations

Usage. On-Line/
Grab
Cation Conductivity CP Discharge
O
Cation Conductivity Polisher Outlet and Economizer O
Inlet
Cation
O
Blowdown or Downcomer
2
Conductivity
Cation Conductivity Hot Reheat Steam or Main
O
Steam
Dissolved Oxygen CP Discharge
O
O
Economizer Inlet
pH (Drum Boilers) Blowdown or Downcomer
O
Sodium
O
CP Discharge
Sodium
Polisher Outlet or Economizer O
Inlet
Sodium
O
Hot Reheat Steam or Main
Steam

Frequency
Measurement
C
C
C
C
C
C
C
C
C
C

Additional Monitoring or Diagnostic Parameters


Parameters

Measurement Locations

pH
Specific
Conductivity
Silica
Phosphate1

Economizer Inlet
Economizer Inlet
Treated Makeup
Treated Makeup
Blowdown or Downcomer

Chloride
Blowdown or Downcomer
Iron
Economizer Inlet
Copper
Economizer Inlet
Total Organic
Condensate Pump Discharge
Carbon
Air In-leakage
Air Removal System
1 Drum Boilers on Phosphate Treatment
2 Drum Boilers on All-Volatile Treatment
O - On-Line
G - Grab

1-8

Usage. On-Line/
Grab

or G

Frequency
Measurement
C
C
C
C
C or S

or G
G
G
G

C or D
W
W
W

C or D
or G
C - Continuous or Semi-Continuous
S - Grab, Once/Shift
D - Grab, Once/Day
W - Grab, Once/Week

EPRI Licensed Material


Introduction

Table 1-2
Generation of Feedwater Corrosion Products by Corrosion and Flow-Accelerated
Corrosion, and the Major Unit Transport and Deposition Problem Areas for All-Ferrous
Systems
Generation

Transport and Deposition

Low/high pressure heater

Deaerator

Boiler deposits and increased boiler pressure


drop

Economizer Inlet

Ripple magnetite formation on the waterwalls


of once-through units

Feedwater Piping

At least five boiler tube failure mechanisms

Drain lines

Frequent need for chemical cleaning

Boiler feedpump fouling

Orifice fouling

Thus for all-ferrous systems, it is clear from this recent research that the optimum
feedwater chemistry involves a transition to more oxidizing conditions (ORP > 0 mV)
and a recognition that oxygen scavengers may not be needed(17). The initial steps
involve gradually eliminating the oxygen scavenger under controlled test conditions,
with the ultimate benefit accruing from the use of oxygenated treatment(14). This
direction is reflected in Sections 3 and 4 of the EPRI AVT Guidelines (15) where it is
suggested that oxygen levels should not be allowed to drop below 1 ppb for all-ferrous
feedwater systems. The optimization of all-ferrous feedwater systems is included in
the Selection and Optimization document(22).
MIXED METALLURGY FEEDWATER SYSTEMS (COPPER CONTAINING)
For mixed metallurgy systems, it is clear that reducing conditions are still required for
the non-ferrous materials (principally copper and copper alloys). The EPRI AVT
Guidelines(15) suggest that oxygen levels should be kept below 5 ppb at the economizer
inlet. However, the type of oxygen scavenger, its level, and the allowable oxygen
concentration that will provide optimal treatment with minimal corrosion product
transport can only be determined by testing.
1-9

EPRI Licensed Material


Introduction

It is difficult and sometimes impossible to control corrosion of both carbon steel and
copper alloys in the feedwater. Control of corrosion of carbon steel requires higher pH
than corrosion control of copper alloys. In copper corrosion, there is a strong
synergistic effect of ammonia plus oxygen and carbon dioxide; ammonium carbonate is
strongly corrosive to copper alloys.
Besides the corrosion of copper alloy feedwater heater and condenser tubing, the main
problem is deposition of copper and its oxides on the inlet stages of high pressure
turbines. This results in rapid loss of generating capacity and turbine efficiency,
requiring chemical and mechanical cleaning of the turbine. Additional problems
caused by copper transport around the water and steam cycle include: deposition in
waterwall tubes and more complicated chemical cleaning and aggravation of corrosion
of cycle components by the deposited copper and its oxides (boiler tubes, turbine
blades, and austenitic stainless steel tubing). Table 1-3 provides an indication of typical
areas suffering from corrosion and flow- accelerated corrosion, and generating
feedwater corrosion products, as well as the major cycle problem areas.
Transport of copper into the turbine occurs both during startups and normal operation.
Feedwater concentrations of copper and its oxides (mostly as colloids) during cold
startup can be as high as 10 ppm. Depending upon the copper alloys used in the cycle,
there may also be elevated concentrations of nickel and zinc.
Table 1-3
Generation of Feedwater Corrosion Products by Corrosion and Flow-Accelerated
Corrosion and the Major Unit Transport and Deposition Problem Areas for Mixed
Metallurgy Systems

Generation

Transport and Deposition

Low/high pressure
heaters, condensers

HP turbine deposits

Boiler deposits and increased boiler pressure drop

At least five boiler tube failure mechanisms

Frequent need for chemical cleaning

Orifice and control valve fouling

Superheater deposits

Heater drain pump fouling

Heater deposits and increased heater pressure drop.

Condenser

1-10

EPRI Licensed Material


Introduction

These feedwater corrosion products either deposit in the boiler or, after concentration
in the boiler water, are carried over into steam as mechanical and vaporous carryover.
Volatility of copper oxides at high boiler drum pressures (>2400 psi) is significant; up
to 30% of boiler water copper can be carried over into the main steam. Because of this
carryover, thousands of pounds of copper can be deposited in superheaters. The
second mode of copper transport into the turbine is by attemperating sprays.
Minimization of the negative effects of copper can be achieved by replacement of
copper alloy tubing (particularly in the HP feedwater heaters), prevention of corrosion
during inactive periods by proper lay-up of heaters and boilers, system filling with
deaerated water, stringent control of air in-leakage, use of condensate polishers
(particularly during startups) and, possibly in the lower pressure units, use of other
amines instead of ammonia. As mentioned previously the choice of the alternate amine
and oxygen scavenger should be determined by comprehensive monitoring.

Alternative Water Treatment Chemicals


During the past 10 years, many new alternate amines and oxygen scavengers have been
introduced. Before applying these chemicals, their properties should be carefully
determined, and experience with each should be verified. Then, within a few weeks of
the first application of the new chemical, the cycle chemistry should be analyzed in
much more detail than during the normal operation.
The selected treatment should protect all cycle components, and prevent generation of
corrosion products, and general and localized corrosion. These chemicals should be
compatible with all cycle component materials and the lay-up practices, and should be
compliant with all environmental and health regulations. Most of the applications are
for low- and medium-pressure industrial units but some of these chemicals are also
being used in utility cycles. The use of some of these products can lead to corrosion,
buildup of deposits, and other problems. It should be kept in mind that the overall
philosophy of the EPRI fossil plant cycle chemistry program is to keep the cycle as pure
as possible with as few chemical additions as possible.
These water treatment chemicals fall into the following categories:

Neutralizing and filming amines for feedwater, steam, and condensate pH control

Oxygen scavengers (hydrazine substitutes)

There are hundreds of formulations of the above chemicals and their mixtures. None of
these chemicals are desirable, needed or recommended in high pressure utility cycles.
However, to evaluate the effects of any water treatment chemical, data pertinent to its
chemical transport, decomposition products, cycle material corrosion, deposit and scale
buildup, and analytical interferences must be known. These needed data include:
1-11

EPRI Licensed Material


Introduction

Hydrothermal stability in the cycle;

Kinetics of reactions;

Decomposition products and their effects;

Analytical interferences;

How to monitor / analyze;

Toxicity of the product, decomposition products, deposits, etc.;

Measured effects on pH, conductivity, cation conductivity, iron and copper


concentration;

Stability in chemical addition tanks and storage containers;

Solubility and volatility of the chemical and its decomposition products;

Behavior of dried-out solutions (deposits in reheaters, superheaters, turbines, valve


gluing); and

Behavior under short-term and long-term lay-up conditions and during startup
(decomposition-acid formation, scale formation, disposal, etc.).

The utility users of water treatment chemicals need to know the pressure and
temperature range of their application and the nature and behavior of the
decomposition products.
GUIDELINE DEVELOPMENTS LEAD TO SELECTION AND OPTIMIZATION
The preceding discussion has provided a synopsis of the EPRI Cycle Chemistry
guidelines that are available, the boiler water and feedwater treatments, and the
problems inherent in all-ferrous and mixed feedwater heater systems:
The situation is that for drum units the following boiler treatments are currently
available:

All-Volatile Treatment (AVT)

Phosphate Treatment (PT), where the Na: PO4 molar ratio is above 2.8, only trisodium phosphate is added, and up to 1 ppm of free hydroxide is allowed

Equilibrium Phosphate Treatment (EPT), where the Na:PO4 molar ratio is above 2.8
at lower levels of PO4 and for cycles with less contamination; the key to these two
phosphate treatments is that phosphate hideout should be minimized

Congruent Phosphate Treatment (CPT), where the Na: PO4 molar ratio in the most
usual version is below 2.6 and both tri-sodium and di-sodium phosphate are added

Oxygenated Treatment (OT), in special cases with high purity water, all-ferrous
feedwater systems, cation conductivity should be <0.15 S/cm for feedwater and
<1.5 S/cm for drum boiler water

1-12

EPRI Licensed Material


Introduction

Caustic Treatment (CT), up to 2 ppm NaOH allowable in the boiler water

For once-through units there are only two choices:

AVT

OT

There is a similar number of feedwater treatments available depending on the system


metallurgy:

Once-through units, all-ferrous, on AVT, with or without oxygen scavenger

Once-through units on OT (all-ferrous)

Drum units, all-ferrous, on AVT with or without oxygen scavengers

Drum units, all-ferrous on OT

Drum units, mixed metallurgy, on AVT with optimum scavenger

In 1997, EPRI published a guidance document which brings together the current state
of knowledge and provides a logical approach to developing the optimum boiler water
and feedwater treatments for each type and combination of unit(22). Clearly the choice
should be based on the system metallurgy, cooling water, historical availability
problems (BTF, turbine blades etc.), historical performance problems, and operating
experience with current chemistry, as determined by comprehensive monitoring
programs.
PROGRAM COPPER
Control of copper corrosion, transport and deposition activity represents a chronic
problem area in some utility plants. Of particular concern is the frequency of turbine
deposition problems which appeared to be increasing in the early and mid 1990s. This
observation, in combination with failures involving flow accelerated corrosion, and
indications that boiler cleaning intervals could not be extended in many cases where
chemistry had been improved, has prompted EPRI to focus more sharply on the science
which governs feedwater corrosion activity and corrosion product transport in mixed
metallurgy cycles.
Examination of the available knowledge base for ferrous and copper alloy materials
employed in the feedwater train of fossil plant units revealed that the understanding
with respect to ferrous materials was satisfactory, at least in a practical sense.
Application of this understanding resulted in development of the latest EPRI guideline
documents covering treatment approaches, which may be used in all-ferrous systems.
Successful application of these approaches has been demonstrated at domestic fossil
plants and at other utilities around the world.

1-13

EPRI Licensed Material


Introduction

The scientific understanding with respect to copper alloys is incomplete, especially


under feedwater heater service conditions. Some areas appear to be in need of basic,
fundamental research and development activity while for others, available information
is deficient, poorly organized and even contradictory. The severity and magnitude of
turbine copper deposition problems were highlighted during a two-day meeting held
in April 1996 in Salt Lake City, Utah. Scientific knowledge and uncertainties with
respect to copper alloys were focused on during the International Conference on
Interaction of Non-Iron Based Materials with Water and Steam, which was organized
by ENEL of Italy, VGB of Germany and EPRI held in June 1996 in Piacenza, Italy(23). In
response to unfavorable trends in the industry and the existence of limitations on
available information, EPRI formulated a strategy to improve the situation, referred to
as Program Copper. It is designed to evaluate and confirm the key features of the
EPRI cycle chemistry guidelines, which pertain to mixed metallurgy cycles.
This State-of-Knowledge document was prepared as the first step in the program. Its
purpose is to summarize the current understanding of copper corrosion, transport and
deposition, to identify areas of deficiency in the knowledge base, and to outline
approaches and strategies which may be pursued by end users dealing with copper
problems. Two other initiatives currently are included in the EPRI response under
Program Copper. The first is research to investigate the volatility and solubility of
copper and copper oxides in superheated steam. The second initiative involves
corrosion studies of typical copper alloys under feedwater chemistry conditions;
particular attention will be given initially to the effect of oxidizing and reducing
feedwater chemistries.
The future direction of Program Copper will consider findings and results of these
initial projects. Both additional laboratory research and development (e.g., copper
corrosion studies) and field investigations (demonstration projects) will be considered
and implemented as necessary to improve control of copper corrosion, transport and
deposition in utility fossil plant cycles.
As noted in other sections of this report, copper transport activity is typically more
significant during unit startups. Development of guidelines for cycling, layup and
startup has been initiated by EPRI for publication in 1998; this should provide further
opportunities for utilities to improve chemistry control related to these conditions.

1-14

EPRI Licensed Material


Introduction

REFERENCES
1. Interim Consensus Guidelines on Fossil Plant Cycle Chemistry. Electric Power Research
Institute, Palo Alto, Calif. CS-4629. June 1986.
2. Monitoring Cycle Water Chemistry in Fossil Plants: Volume 1, Monitoring Results .
Electric Power Research Institute, Palo Alto, Calif. EPRI GS-7556, Vol. 1, October
1991.
3. Monitoring Cycle Water Chemistry in Fossil Plants: Volume 3, Project Conclusions and
Recommendations. Electric Power Research Institute, Palo Alto, Calif. EPRI GS-7556,
Vol. 3, October 1991.
4. Monitoring Cycle Water Chemistry in Fossil Plants : Volume 2, International Water
Treatment Practices in Fossil Fuel Units . Electric Power Research Institute, Palo Alto,
Calif. EPRI GS-7556, Vol. 2, December 1992.
5. Behavior of Ammonium Salts in Steam Cycles. Electric Power Research Institute, Palo
Alto, Calif. EPRI TR-102377. Final Report. December 1993.
6. R.B. Dooley and S. Paterson, Phosphate Treatment: Boiler Tube Failures Lead to
Optimum Treatment. Proceedings: 55th International Water Conference . Engineers
Society of Western Pennsylvania. 1994.
7. R.B. Dooley and W.P. McNaughton, Boiler Tube Failures: Theory and Practice. EPRI
Book TR-105261. 1996.
8. S.L. Goodstine and R.B. Dooley, Behavior of Sodium Phosphates Under Boiler
Conditions. Proceedings: 54th International Water Conference . Engineers Society of
Western Pennsylvania. 1993.
9. Behavior of Sodium Phosphate Under Boiler Conditions. Electric Power Research
Institute, Palo Alto, Calif. TR-102431. June 1994.
10. P. Goldstein and C.L. Burton, A Research Study on Internal Corrosion of High
Pressure Boilers Final Report. Transactions of the ASME, Journal of Engineering for
Power. April 1969.
11. P.R. Tremaine, L.G.S. Gray, B. Wiwchar, P. Taylor and J. Stodola, Phosphate
Interactions with Metal Oxides Under High Performance Boiler Hideout
Conditions. Proceedings: 54th International Water Conference. Engineers Society of
Western Pennsylvania. 1993.
12. Cycle Chemistry Guidelines for Fossil Plants: Phosphate Treatment for Drum Units.
Electric Power Research Institute, Palo Alto, Calif. EPRI TR-103665. Final Report.
December 1994.
13. Cycle Chemistry Corrosion and Deposition: Correction, Prevention and Control. Electric
Power Research Institute, Palo Alto, Calif. TR-103038. Final Report. December
1993.

1-15

EPRI Licensed Material


Introduction

14. Cycle Chemistry Guidelines for Fossil Plants: Oxygenated Treatment. Electric Power
Research Institute, Palo Alto, Calif. EPRI TR-102285. December 1994.
15. Cycle Chemistry Guidelines for Fossil Plants: All-Volatile Treatment. Electric Power
Research Institute, Palo Alto, Calif. EPRI TR-105041. April 1996.
16. Turbine Steam, Chemistry, and Corrosion. Electric Power Research Institute, Palo Alto,
Calif. EPRI TR-103738. August 1994.
17. R.B. Dooley, J. Mathews, R. Pate and J. Taylor, Optimum Chemistry for AllFerrous Feedwater Systems: Why Use an Oxygen Scavenger?. Proceedings: 55th
International Water Conference. Engineers Society of Western Pennsylvania. 1994.
18. Guidelines for Chemical Cleaning of Fossil-Fueled Steam Generating Equipment. Electric
Power Research Institute, Palo Alto, Calif. TR-102401. Final Report. June 1993.
19. Sodium Hydroxide for Conditioning the Boiler Water of Drum-Type Boilers. Electric
Power Research Institute, Palo Alto, Calif. TR-104007. January 1995.
20. Cycle Chemistry Improvement Program. Electric Power Research Institute, Palo Alto,
Calif. TR-106371. April 1997.
21. S.R. Pate, C.E. Taylor, R.C. Turner and T.S. Larson, EPRI Oxygenated Feedwater
Treatment Demonstration Report. Proceedings: 53rd International Water Conference.
Engineers Society of Western Pennsylvania. 1992.
22. Selection and Optimization of Boiler Water and Feedwater Treatments for Fossil Plants.
Electric Power Research Institute, Palo Alto, Calif. TR-105040. March 1997.
23. Interaction of Non-Iron Based Materials with Water and Steam. Edited by: R. B.
Dooley and A. Bursik. Electric Power Research Institute, Palo Alto, Calif. TR108236. July 1997.

1-16

EPRI Licensed Material

2
INTRODUCTION TO COPPER ALLOYS AND UTILITY
INDUSTRY EXPERIENCE

COPPER AND COPPER ALLOYS


Copper and copper alloys have been used since antiquity (ca. 8000 BC) to contain
potable waters, to manufacture instruments and weapons, for coinage, and, over the
past two hundred years, for the fabrication of heat transfer boundaries. The corrosion
resistance was crucial for the use of copper in potable water systems.
Copper is a member of Group IB of the periodic table. Various physical, atomic and
electronic properties of copper, silver and gold (other Group IB metals) are
summarized and compared in Table 2-1.
Table 2-1
Properties of Copper, Silver, and Gold
Cu
29
63.546
0.117
8.92 (20C)

Ag
47
107.868
0.134

Au
79
196.9665
0.134

1083.4C

961.93C

1064.43C

2567C
fcc*

2212C
fcc*

2807C
fcc*

E ( M + / M, V)

7.72
0.522

7.87
0.799

9.22
1.68

E ( M 2+ / M , V)

0.345

1.39

Atomic Number
Atomic Weight
Atomic Radius (nm)
Density (g / cm3 )
Melting
Temperature (C)
Boiling
Temperature (C)
Crystal Structure
Ionization Potential
(eV)

E ( M 3+ / M , V)
Resistivity
(ohm.cm) @ 20C

1.42
1.678 X 10 6

2.24 X 10

1.586X 10

Note: fcc is face centered cubic


2-1

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Three classes of copper rich alloys are of interest: brasses (Cu/Zn), aluminum brasses
(Cu/Al) and cupronickel (Cu/Ni) alloys. These alloys have been used extensively in
thermal power plants in condensers, feedwater heaters and boilers, together with
nickel/copper alloys (Monel). The composition of specific alloys within these classes
are shown in Table 2-2, while Table 2-3 summarizes some of their important properties.
The phase diagrams for the Cu/Zn, Cu/Al and Cu/Ni systems are shown in Figures 212-3.
Table 2-2
Nominal Composition of Commercial Cu-Base Alloys Used in Feedwater Heater
Systems
Common CDA
Name

Number Composition, Weight Percent


Cu

Pb

Fe

Sn

Zn

Ni

Mn

As

Sb

Admiralty
Brass,

0.15T
443

7073

<0.07

<0.06

Arsenical

0.9-

B1

----

----

----

1.2

0.2-

----

1.0

1.8-2.5
687

Arsenical

76.0

<0.07

0.06

----

B1

----

----

----

79.0

Cu-10Ni

706

B1,2

Cu-15Ni

709

Cu-20Ni
Cu-30Ni

0.2-

----

Al
0.15T3

1.0
<0.05 1.0-1.8

----

<1.0

9.0-11.0

1.0

----

----

----

----

1,2

<0.05

<0.6

----

<1.0 13.5-16.5

0.6

----

----

----

----

710

B1,2

<0.05

<1.0

----

<1.0 19.0-23.0

1.0

----

----

----

----

715

B1,2

<0.05 0.4-0.7

----

<1.0 29.0-33.0

1.0

----

----

----

----

Monel 400

30

70

Balance

The sum of the concentrations of elements with specified limits + Cu >99.5

Total other elements

2-2

<0.005
(Al+Si)

Aluminum
Brass,

Other

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Table 2-3
Properties of Some Copper Base Alloys
Alloy Name and
Number

Mechanical Properties

Typical Uses

443 Admiralty brass

Excellent cold working,


fair for hot forming.
Tensile strength 53-90 ksi
(365-660 MPa)

Condenser and
feedwater (LP) heater
tubing

464 -467 Naval brass

Exceptional hot
workability and hot
forgeability. Fabricated
by drawing, bending,
blanking. Tensile
strengths range 5588
ksi.

Condenser tubesheets,
valve stems, marine
hardware, rivets, nuts,
balls, bolts.

687 Aluminum brass

Exceptional cold
workability for bending
and forming. Tensile
strength is 60 ksi.

Evaporator, heat
exchanger, and
condenser tubing,
condenser tubesheets,
distiller tubing.

706 Copper nickel 10%

Decent cold and hot


workability, fabricated
by forming and bending.
Tensile strengths range
from 44 to 60 ksi.

Condenser tubesheets,
and condenser, distiller,
heat exchanger, and
evaporator tubing.

710 Copper nickel, 20%

Decent cold and hot


formability, fabricated by
forming, bending, and
welding. Tensile
strengths range from 49
to 95 ksi.

Condenser tubesheets,
and condenser,
evaporator, and heat
exchanger tubing,
resistors, and ferrules.

715 Copper nickel, 30%

Decent cold and hot


workability, fabricated
by forming and bending.
Tensile strengths range
from 54 to 75 ksi.

Condenser tubesheets,
and condenser, distiller,
heat exchanger, and
evaporator tubing.

2-3

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Zinc is seen in Figure 2-1 to have extensive solubility in copper, and Cu-Zn alloys up to
39wt.%Zn at 450C are -solid solutions. This phase has the face centered cubic (FCC)
structure. The phase diagram for Cu-Al alloys (Figure 2-2) shows that these are
somewhat more complex alloy systems than are the Cu-Zn brasses. Of all the alloys,
the Cu-Ni alloys are the least complex, with copper and nickel exhibiting complete
mutual solubility over the entire composition range. Of course, the addition of a third
alloying element, such as iron, into the copper-nickel alloy can introduce additional
phases.

Figure 2-1 Phase diagram for the Cu-Zn system.

2-4

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Figure 2-2 Phase diagram for the Cu-Al system.

2-5

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Figure 2-3 Phase diagram for the Cu-Ni system.


EARLY UTILITY EXPERIENCE WITH COPPER ALLOYS
Early power plant designs extensively applied copper as the preferred material of
construction for condenser and feedwater heater tubing. As unit capacities and
pressures increased, designers utilized various copper alloys for heat transfer in the
feedwater system. Factors favoring selection of copper and its alloys included high
heat transfer efficiency, low material cost and the availability of technology to
manufacture tubes suitable for the intended service conditions.
Initial designs of fossil units with once-through steam generators employed copper
alloys in the feedwater system. Operation of these units revealed it was essentially
impossible to avoid problems with copper corrosion, transport and deposition activity.
Problems with copper fouling of the turbines of these units in the late 1950s prompted
initial research efforts to appraise the solubility of copper in steam(1). Unfavorable
operating experience with copper alloy heaters in these units resulted in a change in
design philosophy. Subsequent once-through boiler units were equipped with heater
tubes composed of ferrous materials. Many of the copper alloy heaters in the first once2-6

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

through units were replaced with iron-based tubing, although a small number still
operate with tubes composed of Alloy 400 (Monel).
Copper alloys were also used as feedwater heater tube materials in many of the first
domestic nuclear units. Copper and copper oxides were found to be significant
constituents of corrosion product accumulations in the steam generators of these units.
In response to concerns that high crud levels enhanced corrosion activity within the
steam generator, EPRI sponsored research efforts to investigate corrosion of copper
alloys employed in nuclear units. Two studies, the results of which were published in
1982, focused on control of copper alloy corrosion under all-volatile treatment
conditions(2,3).
The EPRI PERSPECTIVE subsection entitled PROJECT RESULTS as published in
Reference 2 is quoted in its entirety as follows:
The release rate of alloy CDA-443 ADM (admiralty brass) was found to be quite high under a
wide range of typical feedtrain environments when operated at a temperature of 193F or higher.
Alloy CDA-706 (90/10 Cu/Ni) was tested at higher temperatures, 274F and 375F, and found
to have similarly high release rates.
A significant fraction of the oxygen from the feedwater reacted with the copper alloys to form
corrosion products. Hydrazine on the other hand reacted with a much smaller fraction of the
oxygen. Based on these test results, Alloy CDA-706 (90/10 Cu/Ni) and Alloy CDA-443 ADM
are not recommended for use in feedwater heaters when there is concern for corrosion product
transport.
Conclusions reached as a result of plant operating experiences and research activities
focusing on nuclear plant concerns resulted in a change in unit design philosophy.
Also, many of the copper alloy heater tube bundles originally installed in these units
were subsequently replaced with copper-free materials.
USE OF COPPER ALLOY HEATERS IN FOSSIL UNITS WITH DRUM BOILERS
Changes in feedwater heater tube material criteria accepted for design of fossil units
with once-through boilers and nuclear units have not been applied as extensively to
designs of fossil units with drum boilers, although recognition of the benefits of
avoiding copper alloys has generally improved over the last 15 years. Replacement of
copper alloy heater tube bundles in fossil units is not normally initiated until justified
by the costs associated with reductions in unit efficiency or reliability as result of heater
performance deterioration.
Today, few utilities continue to replace copper alloy heater and condenser tube
materials without first considering the performance characteristics and costs of
available alternative materials. Unfortunately, the selection is restrained in some
2-7

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

instances because of plant layout and space limitations, this due to the fact that heat
transfer surface area must be increased to provide comparable performance when
replacing copper alloys with copper-free materials. Rearrangement of equipment and
piping to provide the additional space needed is generally cost-prohibitive. In these
cases, it is usually necessary to either install a new heater with tubes of a more
corrosion-resistant copper alloy or replace the heater bundle with the original tube
material.
Drum boiler cycles with mixed metallurgies in the feedwater are still very common,
especially in the case of units constructed prior to 1980; a rough estimate is that there
are between 750 and 1000 units with copper alloys in the feedwater trains in the US.
Continuation of recent trends to extend useful service life of existing units with
minimized capital expenditures implies that a significant amount of domestic power
demand will be supplied by units with copper alloy heaters for the foreseeable future.
PROBLEMS RESULTING FROM USE OF COPPER ALLOYS IN FOSSIL UNITS WITH
DRUM BOILERS
A variety of copper-bearing materials is employed in tubing used in feedwater heaters,
condensers and other heat exchangers (such as gland steam condensers) in the
preboilers or domestic fossil plant drum boiler cycles. A listing of the common
locations of the copper containing materials used is provided in Table 2-4.

Table 2-4
Copper Alloys Employed as Condenser and Feedwater Heater Tubing in Fossil Plant
Cycles
Material
Arsenical Copper
Aluminum Bronze
Aluminum Brass
Admiralty Brass
90-10 Copper-Nickel
80-20 Copper-Nickel
70-30 Copper-Nickel
Monel

Condensers
4
4
4
4
4
-4
--

Low Pressure
Heaters
---4
4
4
4
--

High Pressure
Heaters
-----4
4
4

Figure 2-4 designates locations subject to impurity ingress, corrosion and deposition in
the steam-water cycle of a typical fossil unit with a drum-type boiler. Copper corrosion
may occur at any location - including condenser shellside surfaces, and steamside and
waterside surfaces of feedwater heaters - where steam or condensate is in contact with
copper-bearing materials.
2-8

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Copper ions and/or oxides liberated from the parent material enter the steam-water
cycle. Any line used to transfer condensate or feedwater should be regarded as a
potential transport path for copper. In the case of feedwater heater shellside corrosion
this includes the heater drains and return lines. Copper may enter the steam path as a
result of mechanical and/or vaporous carryover from the steam drum. Here boiler
drum level control is of paramount importance especially during startup. Copper may
also be introduced to the steam path through the attemperation lines to superheaters
and reheaters. The relative importance of copper transport to steam via carryover and
attemperation varies and is dependent on several factors which are discussed further in
Sections 5 and 6.
Blowdown represents the only universally applicable method of removing copper from
the steam-water cycles of drum boiler units. The effectiveness of blowdown as a tool
for controlling copper is rather limited. Further, any benefits derived may or may not
justify the cost of energy losses associated with increased blowdown activity.
Condensate demineralizers and filters are capable of removing much of the copper
released as a result of corrosion in condensers and low pressure feedwater heaters.
Copper released from the high pressure heaters generally cannot be removed by
polishers except in cases where heater drains may be routed to the hotwell or a
feedwater recirculation line is provided; the availability of such features serves to
reduce copper transport during unit startups.
Performance problems may be encountered when deposition of copper occurs before it
can be removed from the cycle by blowdown, polishing or other approaches (e.g.
running heater drains to waste during unit startups). Components which may be
affected include high pressure heaters, boilers and the high pressure section of turbines.
In other instances, copper deposits will increase the level of corrective maintenance
required. Susceptible sites for this include strainers, valves and orifices, all of which
produce a significant pressure drop. Locations between the boiler feed pump and high
pressure turbine are particularly prone to copper deposition activity.

2-9

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Figure 2-4 Typical Locations of Impurity Ingress, Corrosion and Deposition in a Drum
Cycle

Deposition in High Pressure Heaters and on Boiler Waterwalls


Deposition of copper in high pressure feedwater heaters is detrimental in two ways.
Deposit buildup serves to impede heat transfer, which may result in reduced final
feedwater temperatures. A drop of 10F in temperature can result in a one percent
reduction in boiler efficiency. Copper deposits also increase heater pressure drop. As a
result, additional power will be consumed when operating the boiler feed pump to
maintain required flows(4).
Where performance penalties are significant, utilities generally resort to periodic
chemical cleanings to remove the deposits(5,6). Table 2-5 indicates the effect of chemical
cleaning on reduction of pressure drop for two feedwater heaters of a utility fossil plant
unit(7). Cleaning of these heaters removed an estimated 250 pounds (114 kg) of copper.
Performance improvement and reduced power consumption would generally justify a
favorable payback period for the costs associated with cleaning.

2-10

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Table 2-5
Heater Pressure Drop Values Prior to and Following Copper Removal
No. 6 Heater

No. 7 Heater

Before Cleaning

184 psig

175 psig

After Cleaning

120 psig

84 psig

Difference

64 psig

91 psig

Preferential deposition of copper and copper oxides at locations subject to high heat
fluxes is commonly encountered in fossil units experiencing significant corrosion of
copper alloys in the feedwater system. Figure 2-5 A and B are two examples of
laminated deposits consisting of layers rich in copper and magnetite(6). Figure 2-5 A
features accumulations of metallic copper at pit sites in the base metal.
Waterside deposits are of concern since they provide the environment in which any
impurities introduced to the boiler can concentrate to levels at which underdeposit
corrosion activity can be initiated. Damage mechanisms such as hydrogen damage,
caustic corrosion and acid phosphate corrosion can be avoided by simply maintaining
waterwall surfaces in a clean condition(8). Metallic copper, copper oxides and other
copper species, which may be present in waterwall deposits, increase the inventory of
waterside material in which corrosive species can concentrate.
In some instances, waterside deposits have become significant enough to produce
damage and failures related to overheating of the tube metal. Potential effects of
waterside solids accumulations on tube metal temperatures are illustrated in Figure 26(6,9,10). In mixed metallurgy units, the relative proportion of copper in deposits at high
heat flux locations may range from 2050 percent or more.
Again, timely chemical cleaning to remove boiler waterwall deposits is effective in
avoiding underdeposit corrosion activity and preventing solids accumulations from
reaching levels where overheating of tube metal may occur. Extra care must be taken
when planning for chemical cleaning of boilers in units with copper alloy heaters. In
general, significant levels of copper in boiler deposits are likely to require more
frequent cleanings, use of solvents or additives effective for copper removal, longer and
more complicated cleaning procedures, more difficult waste handling and disposal,
and relatively high commitments of time, labor and money to perform the work.
As part of the planning process, it is necessary to estimate the quantity of copper
present in tube deposits which will be removed during the cleaning. In units which do
not use copper alloys for feedwater heaters, these values are usually quite low. For
units with copper alloy heaters, copper removal of 100500 pounds (45230 kg) during
operational cleanings is quite common. Higher recoveries have been reported in larger
units subject to significant feedwater corrosion and transport activity. Removal of more
2-11

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

than 1000 pounds (450 kg) of copper is less common. However, there is one case
reported in the literature where as much as 7508 pounds (3412 kg) of copper was
moved during an operational cleaning of the boiler(11).

Figure 25A Multilayered Deposit with Metallic Copper in Pits

Figure 25B Multilayered Deposit with Metallic Copper Plated on Metal Surface

2-12

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Figure 26 Effect of Internal Scale Thickness on Tube Metal Temperature

Deposition in Turbines
Copper in the steam is likely to deposit at any location subject to pressure losses.
Specific volume increases resulting from the steam passing through the high pressure
turbine makes this area a likely deposition site. The effects of high pressure deposits on
stage efficiency and capacity are staggering. One account indicates that less than two
pounds (1 kg) of copper deposits can reduce turbine capacity by one MW(12). Units
experiencing turbine copper fouling are typically unable to operate for normal time
intervals between major overhauls without either derating the unit or taking
unscheduled outages during which deposits are removed to restore efficiency. High
pressure turbine blade deposits are shown in Figure 27. Many utilities experiencing
turbine performance deterioration have found that conducting steam path audits
during major outages provides an excellent basis for determining and defining the
relative significance of the various factors which may cause efficiency losses(13). Results
of the audits may be used to establish a plan for corrective action(s) which is both
technically sound and cost effective(14).

2-13

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Figure 27 Turbine Copper Deposits


Removal of copperrich turbine deposits may be accomplished in several ways (see
Solution 1 in Section 8). Grit blasting is the most reliable but most costly method,
particularly if an unscheduled outage must be taken to perform the work. Chemical
cleaning with foamed solvents is also generally effective and shortens downtime since
disassembly of the turbine is not required(5, 6, 15). On-line turbine washes, conducted at
reduced pressure conditions which result in moisture formation in the high pressure
turbine have been effective in regaining capacity losses in some but not all instances (16).
More recently, a technique has been developed in which a neutralizing amine is
utilized to soften and displace turbine deposits. This process is based on work
conducted to evaluate processes for reduction of sludge inventories in nuclear steam
generators(17). Field application to fossil plant boilers and turbines is limited and
recovery of capacity losses has been variable(16, 17). Selection of the preferred cleaning
method should consider extent of deposition, deposit characteristics and costs and
benefits associated with removal of the deposits.
Chemical Cleaning is discussed in the Solutions of Section 8.

2-14

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

UTILITY EXPERIENCE AND CONCERNS WITH TURBINE COPPER FOULING


Although significant deposition of copper in utility turbines has been recognized as a
problem since the 1950s, the situation appeared to re-emerge in the 1990s. The
underlying root causes of seemingly more frequent turbine fouling incidents are not
easily explained. Increased utilization of high pressure drum boiler fossil units for
peaking and cycling service may be an important influencing factor, as may be the
increased use of overpressure (5%) operation on the highest pressure units. Certainly
the elimination of the reducing feedwater environment (such as hydrazine), and the
lack of maintenance of the reducing environment during all operating periods have had
major influencing affects. In at least some cases, the time in service (accrued unit
service hours) may have some effect. The relatively recent re-emergence of turbine
copper deposition as an industry issue may be an indication of improved
understanding of this phenomenon on the part of utility personnel tracking unit
performance and cycle chemistry.
The problem should be put in perspective however. Out of the approximately 7501000
units in the US with copper containing alloys less than 100 units have experienced and
recorded serious turbine fouling problems. It is believed that adherence to the
philosophy and limits within the EPRI cycle chemistry guidelines and to the Guiding
Principles discussed later in this Section, will reduce the possibility of fouling. A
significant number of units around the world operate very successfully with copper
alloys in the feedwater system.

Identification of Contributing Factors for Turbine Copper Fouling


The severity and magnitude of turbine copper deposit problems were highlighted
during a two day meeting held in April 1996 in Salt Lake City, Utah. The meeting was
attended by approximately 50 people from about 25 utilities. A survey was conducted
in parallel of 20 utilities which provided data on 45 units, many but not all of which
had experienced capacity reduction and stage efficiency losses as a consequence of
copper turbine fouling. The results of the survey were compiled(16), distributed to the
participants, and were recently published(18).
The scientific knowledge and uncertainties with respect to copper alloys were focused
on during the international conference on the Interaction of Non-Iron Based Materials
with Water and Steam held in Piacenza, Italy in June 1996(19).
Based on these two major resources, plus the survey of over 60 utilities conducted at the
Fifth International Conference on Fossil Plant Cycle Chemistry held in Charlotte, North
Carolina in June 1997(20), a preliminary assessment of the possible Key Contributing
Factors and root causes for HP turbine copper fouling has been developed:

2-15

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Copper fouling becomes significant enough to produce efficiency losses at drum


pressures above 2400 psig (16.5 MPa)

Operating at overpressure of 2750 psig (18.9 MPa) and above exacerbates the
problem further

Both forced circulation and natural circulation boilers can be affected

Unit startups are the period of greatest copper transport activity. Only about 40% of
utilities fill the boiler with deaerated water during startup.

Lack of proper drum boiler water level control, particularly during startup and
transients, can be an important contributor to copper carryover.

The feedwater system is generally not protected during shutdown, i.e. a reducing
environment is not maintained and monitored. Only around 13% of utilities
nitrogen blanket feedwater heaters during shutdown, and only 35% blanket the
boiler. This is considered to be one of the major contributors to the growth and
generation of non-protective copper oxides.

The elimination or reduction of the oxygen scavenger removes the reducing


environment during operation and eventually leads to increased copper transport.
It is very important to note that in some cases it may take several months for this
trend to develop.

Changing the oxygen scavenger also often results in increased transport activity.

Air-inleakage into the LP feedwater circuits increases the growth of non-protective


copper oxides and the copper transport.

The alloys most affected are admiralty brass and 90-10 Cu-Ni in the LP heaters, and
70-30 Cu-Ni and Monel in the HP heaters.

Lack of comprehensive chemistry monitoring and optimization program. For


example over 50% of utilities do not know the iron and copper feedwater levels
during startup.

The survey results(18,20) have also indicated that only about 40% of the respondees
perform any copper monitoring and only 30% monitor iron. Condensate polishing is
only available on about 40% of the affected units; however, copper fouling is
experienced even if the polishers are available and used. Layup practices and
procedures, such as nitrogen blanketing and wet layup with oxygen scavengers, are
used very inconsistently. It also appears that superheating sprays are a less significant
source of copper transport to the turbine than boiler carryover.

Guiding Principles
The EPRI guideline limits for mixed-metallurgy feedwater systems are shown in
Table 2-6.
2-16

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Table 2-6
Normal Cycle Chemistry Limits at the Economizer Inlet for Units with MixedMetallurgy Feedwater Systems
Cycle Chemistry Parameter

AVT Mixed-Metallurgy

pH

8.89.1

Ammonia, NH3 , ppm

0.150.4

Cation Conductivity, S/cm

< 0.2

Fe, ppb

< 5 (<5)*

Cu, ppb

<2 (<2)*

Oxygen, ppb

<5 (<2)*

*Values in parenthesis represent the achievable and desirable limits. These values are achieved
on mixed metallurgy units without any copper problem.

It should be noted that ORP is not included in Table 2-6 as a normal cycle chemistry
limit, however it is considered absolutely necessary to monitor ORP during
customization and optimization of feedwater chemistry control in mixed metallurgy
units. (See Sections 8 and 9).
Based on the contributing factors delineated in the previous subsection, it is considered
that the following Guiding Principles(21), which are inherent in these EPRI guideline
limits, will if followed, lead to successful operation of units with copper alloys:

Keep feedwater copper levels at guideline values (<2 ppb at the economizer inlet)
during normaloperation.

Establish conditions which favor cuprous oxide (Cu2O) rather than cupric oxide
(CuO) under all operating conditions.

Maintain reducing chemistry (oxidizing-reducing potential, ORP < 0 mV) at all


times, including shutdown and startup. This necessitates the use of an oxygen
scavenger.

Control feedwater pH in the range 8.89.1.

Implement shutdown procedures and layup programs which effectively minimize


copper transport activity upon return to service.

Consider volatility effects in controlling drum pressure; if possible, maximize


operating pressure in the range of 24002500 psi (16.517.2 MPa) and avoid
overpressure operation above this range.

2-17

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

The purposes of this document are to bring together the state of knowledge concerning
copper in the fossil plant cycle, and to investigate whether these guiding principles are
adequate to eliminate and/or diminish copper problems in the cycle. Needed further
research and development is identified.
REFERENCES
1. F. J. Pocock and J. F. Stewart. The Solubility of Copper and its Oxides in
Supercritical Steam. Transactions of the ASME, Journal of Engineering for Power.
Vol.
85, Series A, No. 1, January 1963.
2. Effects of Hydrazine and pH on the Corrosion of Copper-Alloy Materials in AVT
Environments with Oxygen.Electric Power Research Institute, Palo Alto, Calif. NP2654. December 1982.
3. Inhibition of Steam Condensate Corrosion of Copper-Based Alloys by Hydrazine
. Electric
Power Research Institute, Palo Alto, Calif. NP-2492. July 1982.
4. Cycle Chemistry Corrosion and Deposition: Corrosion, Prevention and Control.
Electric
Power Research Institute, Palo Alto, Calif. TR-103038. December 1993.
5. Manual on Chemical Cleaning of Fossil-Fueled Steam Generation Equipment
. Electric
Power Research Institute, Palo Alto, Calif. CS-3289. January 1984.
6. Guidelines for Chemical Cleaning of Fossil-Fueled Steam Generation Equipment.
Electric
Power Research Institute, Palo Alto, Calif.. TR-102401. June 1993.
7. J. P. Engle. Chemical Cleaning of Feedwater Heaters. National Association of
Corrosion Engineers, CORROSION/74. Paper No. 104.
8. B. Dooley and W. McNaughton. Boiler Tube Failures: Theory and Practice.
EPRI Book.
TR-105261. 1996.
9. Manual for Investigation and Correction of Boiler Tube Failures.
Electric Power Research
Institute, Palo Alto, Calif. CS-3945. April 1985.
10. American Society For Metals. Metals Handbook, Volume 10, Failure Analysis And
Prevention, 8th Edition.Metals Park, OH, 1976.
11. T. Fitzsimmons, B. Larson, S. J. Arrington and L. Slabodnik. Copper Problems at
Laramie River Station Proceedings: 52nd Annual Meeting, International Water
Conference.Engineers Society of Western Pennsylvania. 1991.
12. G. S. Lawrence. Chemical Cleaning of HP Turbines at Columbia Energy Center.
Proceedings: 56th Annual Meeting, International Water Conference.
Engineers Society of
Western Pennsylvania. 1995.
13. Steam Path Audit Brochure
. Encotech, Inc. 1993.
14. R. Devalois, T. Gilchrist and K. Price. Justification of the Retrofit of a Condensate
Polishing System at Tri-State G&Ts Craig Station. Presented at EPRI Condensate
2-18

EPRI Licensed Material


Introduction to Copper Alloys and Utility Industry Experience

Polishing Workshop; Deep Bed and Powdered Resin Systems.


New Orleans, LA,
September 15-17, 1993.
15. K. J. Shields. Chemical Cleaning of Fossil Plant Equipment. Presented at the 16th
Annual Electric Utility Chemistry Workshop.
Champaign, Illinois, 1996.
16. Turbine Copper Fouling Survey Database,
1996. (Unpublished).
17. Personal communications with Mr. R. Trent, Calgon Corporation, 1997.
18. A. Howell. Review of the 1996 Turbine Copper Fouling Meeting. Fifth
International Conference on Fossil Plant Cycle Chemistry. EPRI Proceedings. TR108459. December 1997.
19. Interaction of Non-Iron Based Materials with Water and Steam
. Edited by B. Dooley and
A. Bursik. EPRI Proceedings. TR-108236. July 1997.
20. Fifth International Conference on Fossil Plant Cycle Chemistry
. EPRI Proceedings. TR108459. December 1997.
21. R. B. Dooley and J. Mathews. The Current State of Knowledge of Cycle Chemistry
for Fossil Plants. Fifth International Conference on Fossil Plant Cycle Chemistry.
EPRI Proceedings TR-108459. December 1997.

2-19

EPRI Licensed Material

3
CORROSION OF COPPER ALLOYS IN FEEDWATER
HEATERS

INTRODUCTION
The concerns related to the transport and deposition of metals and their oxides in the
steam-water cycle of fossil plant units (Tables 1-2 and 1-3) are obviously less significant
when the responsible corrosion mechanisms are effectively controlled. In the feedwater
environment of mixed metallurgy cycles, both carbon steel and copper alloys must be
protected. As shown in Figures 3-1 and 3-2, release rates for iron and copper are pHdependent(1-3). Feedwater pH guidelines established by EPRI for units with mixed
metallurgies (Table 2-6) cover ranges considered the best available for minimizing
release rates for both iron and copper, yet optimal for neither(46). Optimization of these
guidelines for most effective use in individual fossil plant units requires a continual
program of chemistry surveillance, including monitoring of metals transport under
typical operating conditions(7).
As was indicated in Table 2-2, a variety of copper alloys has been used for the tube
bundles of feedwater heaters. Brasses are typically used in condenser applications but
are also sometimes used in low pressure heaters. The copper-nickel alloys are
employed in both condenser (particularly in the air removal section and areas requiring
better erosion-corrosion resistance than brass) and feedwater heater service. Alloy 400
(Monel) is only used in high pressure heaters.
Field experience has shown copper alloys used in feedwater heaters and condensers to
be prone to corrosion by several mechanisms. As indicated in Table 3-1, some of these
mechanisms result in tube failures with relatively minor copper release while others
can allow significant release of copper before failures are experienced(8-10). Corrosion
mechanisms which typically result in tube failures with relatively low release of copper
generally involve deficiencies in the design and/or operation of the heater. Discussion
of these mechanisms, their root causes, and appropriate investigative and corrective
actions is available elsewhere(8,9,11).
General corrosion is typically regarded as the mechanism of greatest concern in units
with properly designed and operated copper alloy heaters. Release of copper via
exfoliation also is significant in those units containing susceptible copper-nickel alloys,
especially when subjected to cycling and peaking service. These mechanisms account
3-1

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

for most of the copper present in boiler feedwater when units are operated under
guideline chemistry conditions (Table 2-6). In the case of marginal heater designs,
copper release associated with erosion-corrosion activity also may be significant,
particularly if the number of tubes affected is high.
Dealloying - which preferentially dissolves zinc from brass or nickel from coppernickel alloys, leaving behind a spongy copper surface which has minimal corrosion
resistance - also has been cited as a means by which copper release from feedwater
heaters may occur. Two separate metallurgical investigations related to a utility
turbine copper fouling problem concluded that dealloying of 70-30 copper-nickel
feedwater heaters was a significant part of the copper release mechanism(12).
Dealloying of copper alloys is more commonly encountered on the waterside surfaces
of tubes used in condenser applications(1315).

Source: Adapted from References 1 and 2

Figure 3-1 Average Copper Release as a Function of Feedwater pH for Various Copper
Alloys

3-2

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Source: Reference 3

Figure 3-2 Corrosion Product Release in ppb for Carbon Steels as a Function of
Feedwater pH

3-3

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Table 3-1
Corrosion Mechanisms Influencing Copper Alloys in Feedwater Heater Service
(a) Water Side, and (b) Steam Side
Water Side Mechanisms

Materials
Admiralty
Brass

90-10
Cu-Ni

85-15
Cu-Ni

80-20
Cu-Ni

70-30
Cu-Ni

Monel

General Corrosion1

Stress Corrosion Cracking2

Erosion-Corrosion2

Steam Side Mechanisms

Materials
Admiralty
Brass

90-10
Cu-Ni

85-15
Cu-Ni

80-20
Cu-Ni

70-30
Cu-Ni

Monel

Condensate Grooving2

Stress Corrosion Cracking2

Acid Attack2,3

Exfoliation1

Impingement2

Fretting2

S = Susceptible
R = Resistant

P = Potentially Susceptible Under Severe Conditions


N = Not Applicable to Condenser/Low Pressure Heater
Applications
Note 1: This mechanism can allow high release of copper, usually without producing tube failures.
Note 2: This mechanism typically causes tube failures with relatively low release of copper.
Note 3: This mechanism only occurs when sulfite is used as an oxygen scavenger.

COPPER OXIDE FORMATION AND STABILITY


Copper forms two common oxides (Cu2O and CuO) and one common hydroxide
(Cu(OH)2) in aqueous feedwater environments. Cuprous oxide, Cu2O, is a well known
p-type semiconductor, which is consistent with the predominant defect being the
3-4

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

cuprous vacancy ( VCu


). Cupric oxide, CuO, is reported to display n-type
semiconductor characteristics, presumably due to the existence of oxygen vacancies
( VO ) as the principal defect. The defect character is particularly important in
analyzing the mechanism(s) of growth and breakdown of passive films. Numerous
studies have shown that the passive film on copper (alloys) is a bilayer structure
comprising an inner defective Cu2O layer that grows into the metal, and an outer
Cu(OH)2 or CuO layer that incorporates species from the feedwater solution.
The oxides and hydroxide of copper are well characterized compounds, and some of
their properties are summarized in Table 3-2.

Table 3-2
Chemical, Structural, and Thermodynamic Properties of the Oxides and
Hydroxide of Copper
Species

Molecular
Weight

Density
(g/cm 3 )

Structur
e

Principal
Defect

o
G f
(kJ/mol)

H f
(kJ/mol)

(J/K.mol)

(J/K.mol)

Cp

Cu 2 O

143.08

6.104

Cubic

VCu

-146.02

-168.62

93.14

63.64

CuO

79.54

6.508

Monoclinic

VCu
and

-129.70

-157.32

42.63

42.3

Crystalline

None

-359.02

-443.09

87.03

87.85

Cu(OH )2

97.56

3.368

/or VO

The oxide formation process on copper alloys in feedwater environments initiates with
the oxidation of metallic copper to cuprous oxide by the following possible reactions.
2 Cu + O2 2 Cu2O

(3-1)

2 Cu + H2O Cu2O + H2

(3-2)

Reaction (3-1) is thermodynamically favorable at 25C; the thermodynamic feasibility of


reaction (3-2) is questioned by some investigators(1517) and is rarely cited as the
predominant corrosion reaction.
Cuprous oxide formed as a result of reactions (3-1) and (3-2) subsequently may be
further oxidized to cupric oxide.

3-5

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Cu2O + O2 2 CuO

(3-3)

or Cu2O + H2O 2 CuO + H2

(3-4)

Again reaction (3-4) is questioned by some workers(1517) and is rarely cited as the
predominant corrosion reaction. It is appropriate to consider the possibility that cupric
hydroxide (Cu(OH)2), another valence 2 compound, will form instead of cupric oxide.
The conversion of Cu(OH)2 to CuO can be described by the equilibrium reaction
Cu(OH)2 CuO + H2O

(3-5)

At the temperatures of interest (<200C, 390F), the standard Gibbs energy (RG) is
negative, showing that CuO is stable relative to Cu(OH)2. Thus at no feedwater
temperature of interest is Cu(OH)2 stable relative to CuO. This is important because it
implies that any Cu(OH)2 that is observed in contact with the feedwater cannot form by
the hydration of CuO, but must form from a less stable species (such as Cu2+, Cu+, or
possibly Cu2O, depending on the specific conditions).
When oxygen concentrations are high enough, both Cu2O and CuO will form on a
copper or copper alloy surface. As shown in Figure 3-3, cuprous oxide is the
predominant species at the metal/oxide interface. The proportion of cupric oxide
present increases with distance from the base metal(18). For feedwater conditions within
the Guiding Principles (Section 2), (reducing conditions, low oxygen, oxygen
scavenger/reducing agent, proper pH range) the oxide layer has been observed to
consist mainly of cuprous oxide. However, when the Guiding Principles are broken by
allowing oxidizing conditions (ORP > 0 mV) to exist by eliminating the oxygen
scavenger or because of air-inleakage, then the proportion of cupric oxide in the outer
layers increases. There is a concomitant increase in the amount of porosity, which
facilitates the mechanical disruption and release of the CuO into the feedwater.
In general, the surface oxide layer found on copper and its alloys tends to be less stable,
less passive, and less protective than the iron oxide films found on ferrous alloys(19).
Destabilization of the copper oxide film involves both chemical and
physical/mechanical processes. Both cuprous and cupric oxides are subject to
dissolution, with higher solubility noted for the latter, at least under reducing chemical
conditions. However, as already noted, CuO predominates under oxidizing conditions
and the solubility of CuO under such conditions appears to be lower than the solubility
of Cu2O under reducing conditions, according to one study(51) and in accord with the
known thermodynamic properties. Nevertheless, solubility does not necessarily
control release rate, so care is needed in interpreting this information. Mechanical
strain present in the oxide layer renders it subject to fracturing as distance from the
3-6

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

parent metal increases. As a consequence of these processes, the diffusion rate of


copper increases, leading to higher release rates. Porous oxide layers will be subject to
high rates of corrosion, while tighter layers will restrain corrosion and release of copper
to the feedwater.

Source: Reference 18

Figure 3-3 Schematic Cross-sectional Diagram of the Structure and Charge Mobility
within a Copper Oxide Layer Formed on Pure Copper Metal in Aqueous Solution
Typically, approaches to minimize copper corrosion and release have focused on
control of those species known to promote dissolution of metallic copper. Under
3-7

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

reducing chemistry conditions, the process is generally expected to consist of two steps,
oxidation of copper and production of copper ions. Dissolved oxygen in the water
serves as the oxidant while ammonia and, perhaps, certain neutralizing amines, serve
to form stable complexes with copper ions.
This concept is supported by research which demonstrates that the principal cathodic
reaction for copper alloys in solutions containing ammonia is cupric ion reduction(21).,
although dissolved oxygen or ferric iron will also serve as cathodic depolarizers. It
should be noted, however, that the presence of ferric ion and cupric ion in mixed
metallurgy systems invariably implies the presence of oxygen, although the possibility
that hydrogen evolution becomes a viable cathodic reaction must also be recognized.
Initially, in the absence of cupric ion, copper corrodes at a slow rate that is governed by
the kinetics of oxygen reduction. The cuprous ion in solution is then complexed by
ammonia to form Cu(NH3)2+ which, in turn, reacts with oxygen to form the cupric
complex ion according to the reaction:
2 Cu(NH3)2+ + O2 + H2O + 4NH3 2 Cu(NH3)4++ + 2 OH-

(3-6)

This reaction of oxygen with the cuprous complex ion in the bulk occurs much more
rapidly than reaction of oxygen at the metal surface because the bulk reaction is
homogeneous in nature (i.e., it occurs uniformly throughout the solution) while the
reaction at the metal surface is heterogeneous. When sufficient quantities of the cupric
complex ion are generated, they become the primary oxidizing species and rapidly
accelerate corrosion of susceptible copper-base alloys with the concomitant production
of cuprous complex ions, Cu(NH3)2+. These reduced species are then regenerated by
oxygen in the solution and thus the dissolution process is autocatalytic in the presence
of oxygen.
FACTORS INFLUENCING COPPER CORROSION AND RELEASE ACTIVITY
As indicated above, corrosion of copper and its alloys is strongly influenced by the
presence of dissolved oxygen and ammonia in solution. However, the body of research
available indicates that several other factors are also important. The interaction of all
these factors is not fully understood.
The terms corrosion rate and release rate are often used interchangeably in
discussions of metals transport. In most contexts, this is permissible. However, when
considering the root causes of copper corrosion and release, and the influencing factors
which either inhibit or enhance the process, researchers generally choose to make one
important distinction. Release rate includes entry of copper into the feedwater by
both dissolution (as ions) and as a result of physical breakdown and release of the
oxide layer (as colloidal or suspended solids). Table 3-3 indicates some forms of copper
3-8

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

that may be present in feedwater. Other dissolved and solid species derived from the
alloying elements (Ni, Zn, etc) are also often detected. Corrosion rate not only
includes release of dissolved and suspended copper species into the feedwater but also
reflects any metal oxidized in establishing or maintaining the surface film(22).
The subsections which follow address the various factors which influence copper alloy
corrosion and release rates. Both laboratory findings and field observations are
presented. Some of the data presented reflect temperatures that exist in high pressure
heaters but much of it applies to condenser and low pressure heater temperatures. In
general, the overall effect of varying an individual influencing factor is readily
definable, however, the effect on individual processes (film formation, metal
dissolution, release of suspended solids) is more difficult to establish. Interpretation
becomes more complex when two or more influencing factors are varied, as is normally
the case in the field. Seemingly synergistic or antagonistic effects observed under such
conditions may involve other variables, the influence of which are either not recognized
or incorrectly assumed to be constant. Also, the relatively short time frame over which
laboratory experiments or field investigations are normally conducted may not be
reliable if extrapolated over the entire useful service life of a feedwater heater.

Table 3-3
Forms of Copper which may be Present in Feedwater
Solids

Oxides: Cu2O, CuO

Hydroxides: Cu(OH)2

Other species (such as chlorides and sulfates) may form if the water is contaminated.

Dissolved Species

Cations: Cu+, Cu++, Cu(NH3)2+, Cu(NH3)4++

Anions: HCuO2-, CuO2--

Other species (such as copper-chloride complexes) may form if the water is contaminated.

Despite these uncertainties, results of technical investigations have been used, in


combination with field experience and observations, to establish effective and
attainable chemistry guidelines for metals in the feedwater of fossil units. Field
chemistry monitoring sponsored by EPRI indicated that the guidelines for feedwater
copper (Table 2-6) could be maintained during normal operation of a drum boiler unit

3-9

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

(Figure 3-4). Surprisingly, slightly elevated copper levels were determined for a drum
boiler unit with a condensate polisher and copper alloy condenser tubes (Figure 3-5).

Source: Reference 23

Figure 3-4 Average Copper Corrosion Product Transport Rates for Normal Operation
at Fox Lake Unit 3

Source: Reference 23

Figure 3-5 Corrosion Products at the Economizer Inlet and in the Reheat Steam for
Typical Operation at Newton Unit 1

3-10

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Survey data for the latter unit include no mention of turbine fouling or other problems
with copper, despite the fact that the boiler drum pressure was reported to be 26202800 psig(23). International utilities also generally try to limit feedwater copper levels to
1-3 ppb (with actual readings typically 1 ppb) and report no problems related to
turbine copper fouling when the EPRI criteria are attained(24).

Effect of Oxygen
As indicated by Syrett in a recent paper reviewing corrosion of copper alloys(22), most
published accounts of laboratory investigations and field research indicate increasing
rates of metal release from brasses and copper-nickel alloys as oxygen concentration is
increased from <30 ppb to 100500 ppb. Graphical representation of results of this type
are depicted in Figure 3-6(25, 26). In these experiments, selected copper alloys were
exposed to neutral water with either a low (3 ppb) or high (100200 ppb) concentration
of dissolved oxygen at a moderate flow rate (5 ft/sec, 1.52 m/sec). Metal release rates
for each alloy are plotted against temperature for the low and high oxygen waters.
Unlike the brasses and copper-nickel alloys, copper release rates for Monel are
relatively insensitive to oxygen level, probably because of the high nickel content,
which leads to the formation of NiO passivating film. Similar trends have been
observed by other investigators including Burgmayer(27), Sato(28), and Castle et al(29).
The effect of increased dissolved oxygen levels on copper alloys also was found to be
detrimental in a boiling water reactor system(30), a drum boiler fossil unit(31) and a
Russian fossil unit with a once-through supercritical boiler when converted from
deoxygenated AVT to oxygenated treatment (OT)(32).
A small, but potentially important, body of work has been published which indicates
that increased oxygen levels may reduce the rates of copper release under some
conditions(33,38). For instance, a 1980 paper(33) indicated that it should be possible to
operate mixed metallurgy cycles under a combined water treatment regime with
oxygen levels ranging from 150300 ppb. Field evaluations of copper release from
units maintained under oxidizing environments (ORP > 0 mV) in Australia(34),
Denmark(35) and Italy(36) have shown this approach to be less than ideal. High rates of
copper transport resulted in the Australian unit returning to reducing (hydrazine)
treatment. Turbine copper fouling was experienced in Denmark. Investigations in
Italy are continuing, but feedwater copper levels measured under oxygenated
treatment are higher (45 ppb Cu when using hydrogen peroxide, 2.53.5 ppb Cu when
using oxygen) than when utilizing AVT (reducing) chemistry (0.51.5 ppb Cu).
The various research studies and field investigations show that oxygen is an important
but not necessarily dominant factor in the corrosion of copper alloys. However, they
also indicate the complexity of the situation and identify other factors which will affect
rates of copper corrosion and release under fossil plant operating conditions. In a
properly operated unit, these would appear to include pH (or concentration of additive
applied for pH control), presence and concentration of oxygen scavenger, service
3-11

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

temperature, and feedwater velocity. The effect of a selected feedwater condition will
be different for each copper alloy to which it is applied. Corrosion/release rates would
be affected further during periods of dissolved solids contamination such as can occur
when the condenser leaks.
Oxygen scavengers appear to be beneficial since they react with dissolved oxygen and
establish a reducing environment which has been shown to reduce copper
corrosion/release rates under some laboratory and field conditions. Testing performed
in Australia to evaluate metals transport in feedwater treated with oxygen, hydrazine
and ammonia concluded that copper corrosion and/or transport could not be
effectively controlled under oxidizing conditions. As shown in Figure 3-7, increased
hydrazine residuals were beneficial in controlling copper release under reducing
conditions created with conventional volatile treatment at a pH of 8.80(34).
A negative aspect of proprietary scavenger treatments is that they, like neutralizing
amines, are subject to thermal breakdown in high temperature areas of the cycle.
Generic hydrazine, still widely used and regarded as the scavenger of choice where
permitted, breaks down to ammonia. Degradation of propriety oxygen scavengers not
only represents a possible source of ammonia, but also generates an inventory of
organic species which ultimately are converted to stable organic acids (formic, acetic)
and carbon dioxide, which can increase the cation conductivity of steam markedly.

[Note: approx. release rates in units of m/y can be obtained by multiplying rates in g/m2.y by 0.11]
Source: Reference 22

Figure 3-6 Steady State Metal Release Rates in Neutral Water Flowing at 5 ft/s
Containing (a) Low Oxygen (3 g/kg O2 with 100200 g/kg H2) and (b) High Oxygen
(200 g/kg O2 with 25 g/kg H2). [From refs. 25 and 26.]

3-12

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Source: Reference 34

Figure 3-7 Effect of Hydrazine Concentration on Feedwater Copper Transport at pH


8.8 in an Australian Unit

Effect of pH
Feedwater pH is controlled by addition of ammonia or neutralizing amines. Other
species which, if present, will influence feedwater pH include carbon dioxide (present
as a consequence of condenser air inleakage), oxygen scavengers, and organics (present
in the treated makeup or as breakdown products of proprietary treatment chemicals).
Laboratory investigations of copper alloy corrosion typically control pH with ammonia
or an amine. The other influencing species are either assessed individually or excluded
entirely. The opposite situation exists in the power plant, where all species may be
present but are less effectively controlled and may not be measured.
As indicated in Figures 3-8 and 3-9, corrosion/release rates for copper alloys appear to
decrease and then increase over the pH range of 510(37, 38). The pH at which corrosion
is minimum appears to fall between pH values of 8.0 and 9.0, but is sensitive to
temperature and possibly to other factors. Other investigations have suggested
corrosion rates to be less strongly influenced by pH, as shown in Figure 3-10(39, 40). One
published account indicated minimum copper corrosion may occur at higher pH
values(41) when the oxygen concentration is high (1 atm. oxygen). These investigations
focused primarily on condenser and low pressure heater service conditions.
Copper release rates for feedwater treated with neutralizing amines would probably be
different than for feedwater dosed with ammonia due to presumed differences in
3-13

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

copper complexing abilities, which are not well defined. Copper-cyclohexylamine


complexes tend to precipitate from solution which may form a protective film on the
base metal. Laboratory evaluations have shown that copper release rates increase
following a change from cyclohexylamine to ammonia; whereas the release remained
stable when switching from ammonia to cyclohexylamine(27).
Use of volatile neutralizing amines instead of ammonia to control condensate and
feedwater pH is sometimes implemented to control corrosion of copper alloys. This
strategy has failed in virtually every case where no effort was made to improve
dissolved oxygen control. Although some amines exhibit better thermal stability than
others, in-service degradation of these treatments will produce some ammonia in
addition to other breakdown products, including carbon dioxide and organic acids.
Another concern is the tendency of some amines to liberate preexisting deposits. As a
precaution, system cleanliness should always be evaluated as part of efforts to consider
any proposed change in feedwater treatment.

Source: Reference 22

Figure 3-8 Effect of pH on Copper Release Rate From 71Cu/28Zn/Sn (SoMs-71 brass,
or Admiralty brass) Exposed to Water Containing Carbon Dioxide (for pH <7) or
ammonia (for pH>7) at Temperatures of 30C and 90C. [From Ref. 37]

3-14

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Source: Reference 22

Figure 3-9 Effect of pH on Steady State Release Rates for 90Cu/10Ni and 70Cu/30Ni
Exposed to Ammonia Solutions Containing 8 - 21 g/kg Oxygen, Flowing at ~ 1 ft/s
(0.3 m/s) and at a Temperature of 35 - 38C. [From Ref. 38]

3-15

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Source: Reference 22

Figure 3-10 Corrosion Rates for Copper Alloys Exposed to Ammonia Solutions
Containing 100 200 g/kg Oxygen* at 40C. pH 7, 9.4 and 10 Solutions were
Obtained by Adding 0, 2, and 20 mg/kg NH3 Respectively. [From Refs. 39 and 40]
[*Note: The quoted oxygen concentration may be in error: the sparging gas composition
reported in ref. 39 suggests that the dissolved oxygen concentration should be
<1g/kg.]

Early research into condensate grooving damage of copper-alloy condenser tubes


determined copper release in ammoniated, oxygen-rich condensate was accelerated if
carbon dioxide, carbonate or bicarbonate was present. Results of one such assessment
are shown in Table 3-4(42).

3-16

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Table 3-4
The Influence of Addition of 400 ppm (NH4)2CO3 on the Corrosion Rate of Selected
Copper-Based Alloys in an Ammoniacal Solution (500 ppm NH3) at Ambient
Temperature
Alloy

Corrosion Rate, mpy


With (NH4)2CO3

Without (NH4)2CO3

Arsenical Copper

95

3.0

Admiralty Brass

16

6.5

Aluminum Brass

8.0

2.0

Aluminum Bronze

13

1.0

90-10 Copper Nickel

5.0

1.0

70-30 Copper Nickel

4.5

0.5

Source: Reference 42

The exact role of carbon dioxide in copper corrosion is not fully understood. If all other
factors are held constant, elevated carbon dioxide levels will require higher ammonia
or amine treatment levels to maintain feedwater pH. Organic acids would exhibit
similar influence over feedwater treatment requirements. In either case, the level of
complexing agent needed to maintain pH would increase. This effect alone appears to
be significant. Carbon dioxide may also accelerate the corrosion by depolarizing the
cathodic reaction(52), or there may exist at elevated temperatures yet unidentified
carbonate complexes of the form Cu( HCO3 )

2 n
2 2n
and/or Cu HCO3
.
n

Effect of Velocity
This parameter is considered in relatively few evaluations of copper alloy corrosion
and release rates. Figure 3-11 presents results of one such study which determined that
velocity effects were more pronounced in aerated water then in deaerated water(43).
Test velocities were 3, 6.5 and 10 m/s, all greater velocities than normally applied when
designing heaters for the alloy evaluated, admiralty brass.
As previously mentioned, assessment of copper release from 90-10 copper-nickel(27)
revealed little significant change when the flow rate was increased from 0.3 ft/s (0.091
m/s) to 4.5 ft/s (1.37 m/s) in deaerated (oxygen <13 g/kg water. Result for parallel
3-17

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

tests with aerated water are considered inconclusive due to the variable oxygen levels
used during this phase of the program. The velocities examined were lower than those
typically used in heater design.
These limited data appear to support the view that corrosion rates are controlled by
surface film solubility and particulate release rates when oxygen levels are low. With
elevated oxygen concentrations, mass transport effects appear to become more
significant; this may include either transport of oxidants (oxygen, cupric ammonium
complex ions) to the surface or of reduced species (copper ions or cuprous ammonium
complex ions) away from the surface and into the bulk solution.
A review of erosion-corrosion of copper alloys by Macdonald(44) indicates that most
attention has been given to improved understanding of inlet end damage by circulating
water entering condenser tubes. These evaluations have indicated that surface shear
stresses may play some role in disruption of the oxide film at high velocity regions(45).
However, other investigations concluded such effects to be minimal, instead
emphasizing the importance of local electrochemistry and mass transfer conditions
which are more influenced by turbulence intensity than shear stress(4648).
Specifically(44,47), it was found that the corrosion potential shifted in the positive
direction with increasing flow velocity, which is consistent with cathodic (oxygen
reduction) control over the corrosion process. At a sufficiently high flow velocity (the
critical velocity) the corrosion potential is approximately equal to that observed for
passivity breakdown on copper in the absence of flow. It is for this reason that the
primary reason for flow-induced corrosion of copper in oxygenated seawater was
attributed to electrochemical factors (shift in the corrosion potential) rather than to
mechanical phenomena (flow-induced shear). However, it is likely that a whole
spectrum of mechanisms exists from essentially mechanically-induced breakdown to
electrochemically-induced breakdown, when flow-induced effects are considered in the
corrosion of power plant materials. To date, these effects have been poorly defined,
principally because flow-induced corrosion (erosion-corrosion) was previously
interpreted in terms of mechanical effects (surface shear) along with electrochemical
factors being totally ignored. The mechanical view became largely untenable when
Syrett(46) pointed out that the surface shear stress generated at the critical velocity
appeared to be far too small to explain the onset of erosion-corrosion in copper and
copper alloys.

3-18

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Source: Syrett, Reference 22

Figure 3-11 Effects of Velocity and Deaeration on the Reduction of the Wall Thickness
of Admiralty Brass Tubes at 185C. (Water was aerated for the first 2200 hours of
exposure and deaerated thereafter.) [From Ref. 43]
Maximum feedwater heater tube side velocities for various materials approved by the
Heat Exchange Institute are shown in Table 3-5(49). These values are actually arithmetic
average values for the entire tube bundle which are used for design and performance
testing purposes. Therefore, actual velocities in certain tubes of a heater designed in
accordance with these standards may be greater than the maximum feedwater
velocities indicated in the table. For further discussion of tube velocity see Section 8,
Solution 8 and Table 8-5. Plugging of leaking or thinned heater tubes during
maintenance outages further increases feedwater velocities relative to design criteria.

3-19

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Table 3-5
Heat Exchange Institute Criteria for Maximum Feedwater Heater Tube Velocities
Material

Tube Velocity, ft/s

Stainless Steel, Monel

10.0

Copper-Nickel Alloys

9.0

Admiralty Brass and Copper

8.5

Carbon Steel

8.0

Effect of Temperature
Generally, rates of copper alloy corrosion and release increase as temperature is
elevated. This effect was illustrated earlier in Figure 3-6. A Japanese study of various
copper-nickel alloys (Figure 3-12) indicated similar results, with the influence of
temperature becoming relatively minor as nickel levels increased(28). The latter data are
generally consistent with those in Figure 3-6(b) for the copper-nickel alloys (10%, 30%
Ni) in oxygenated water, but appear inconsistent with the data for Monel (70% Ni). For
this alloy, temperature does have a significant effect in Figure 3-6(b) but does not in
Figure 3-12. It is possible that the reason for the difference lies in the parameter used to
measure attack: descaled weight loss in Figure 3-12 and release rate in Figure 3-6, i.e. as
temperature increases; the rate of surface film formation may decrease at about the
same rate as the release rate increases.
Temperature also appears to be significant in corrosion and release involving
exfoliation activity(50). Field experience has shown that the susceptibility of
cupronickels to exfoliation increases with nickel content. This trend apparently does
not continue indefinitely since Monel has been shown to be extremely resistant to
exfoliation in the power plant environment. Brasses have been shown to be subject to
exfoliation under laboratory conditions but this mechanism has not been observed in
the field. The experimental work on brasses was conducted at 300C, which is well
above temperatures present in low pressure heaters.

3-20

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

Source: Reference 22

Figure 3-12 Effect of Nickel Content (0-100%) and Temperature (200 to 350C) on the
Corrosion Rate of Copper-Nickel Alloys Exposed for 250 Hours to Oxygen-Bearing
Water Under Saturated Pressure Conditions. [From Ref. 28.]

CONCLUDING REMARKS ON THE CORROSION OF COPPER ALLOYS


The Guiding Principles delineated in Section 2 and the philosophy of the EPRI
guideline limits (Table 2-6) for mixed-metallurgy systems are currently directed
towards the production and maintenance of predominantly Cu2O films under reducing
feedwater conditions. However, there is no consistent quantitative understanding of
the effects of the key variables on the corrosion of copper alloys in terms of the oxide
layers formed. Thus there needs to be confirmation, on a priority basis, of the
predominant oxides formed in both oxidizing and reducing feedwater under as
realistic conditions as possible. There also needs to be a clear definition of the effects of
pH, ammonia, oxygen, temperature, velocity, and of impurities (mainly carbon
dioxide) in the cycle.

3-21

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

REFERENCES
1. Corrosion-Product Transport in PWR Secondary Systems , Electric Power Research
Institute, Palo Alto, Calif. EPRI NP-2149. December 1981.
2. Effects of Hydrazine and pH on the Corrosion of Copper-Alloy Materials in AVT
Environments with Oxygen. Electric Power Research Institute, Palo Alto, Calif. EPRI
NP-2654. December 1982.
3. F. Gabrielli, How to Protect Boilers and Auxiliary Equipment During Lay-up and
Cyclic Service. Presented at the Electric Utility Workshop on Water Treatment and
Corrosion Control. Champaign, Illinois, April 1985.
4. Cycle Chemistry Guidelines for Fossil Plants: All-Volatile Treatment. Electric Power
Research Institute, Palo Alto, Calif. TR-105041. April 1996.
5. Cycle Chemistry Guidelines for Fossil Plants: Phosphate Treatment for Drum Boilers.
Electric Power Research Institute, Palo Alto, Calif. TR-103665. December 1994.
6. Sodium Hydroxide for Conditioning the Boiler Water of Drum-Type Boilers . Electric
Power Research Institute, Palo Alto, Calif. TR-104007. January 1995.
7. Selection and Optimization of Boiler Water and Feedwater Treatments for Fossil Plants .
Electric Power Research Institute, Palo Alto, Calif. TR-105040.
8. Symposium on State-of-the-Art Feedwater Heater Technology . Electric Power Research
Institute, Palo Alto, Calif. EPRI CS/NP-3743. October 1984.
9. Recommended Guidelines for the Operation and Maintenance of Feedwater Heaters .
Electric Power Research Institute, Palo Alto, Calif. EPRI CS-3239. September 1983.
10. K. J. Shields. Corrosion, Transport and Deposition of Copper in Fossil Utility
Cycles. Presented at the Southern Company Services Water Chemistry Seminar,
1996.
11. High-Reliability Feedwater Heater Study . Electric Power Research Institute, Palo Alto,
Calif. EPRI CS-5856. June 1988.
12. G. S. Lawrence, Chemical Cleaning of HP Turbines at Columbia Energy Center.
Proceedings: 56th Annual Meeting, International Water Conference . Engineers Society of
Western Pennsylvania. 1995.
13. Corrosion-Related Failures in Power Plant Condensers . Electric Power Research
Institute, Palo, Alto, Calif. EPRI NP-1468, TPS 79-730. Final Report, August 1980.
14. Seminar Proceedings: Prevention of Condenser Failures - The State of the Art . Electric
Power Research Institute, Palo Alto, Calif. EPRI CS-4329-SR. December 1985.
15. T. E. Erikson, P. Ndalamba, and I. Grenthe, On the Corrosion of Copper in Pure
Water, Corrosion Science, Vol. 29, No. 10, p. 1241, 1989.
16. G. Hultquist, Corrosion Science, Vol. 26, p. 173, 1986.
3-22

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

17. G. Hultquist, G. K. Chuah, and K. L. Tan, Comments on Hydrogen Evolution from


the Corrosion of Pure Copper, Corrosion Science, Vol. 29, No. 11/12, p. 1371, 1989.
18. P. R. Burgmayer and A. C. McDonald. The Corrosion and Release of Copper in
Steam Generation Systems Presented at the Edison Electric Institute, October 2-4,
1988. Betz Technical paper 307.
19. H. Leidheiser. The Corrosion of Copper, Tin and Their Alloys . Robert E. Krieger
Publishing Company, Huntington, NY, 1979.
20. D. D. Macdonald. Condensate Corrosion. Seminar Proceedings: Prevention of
Condenser Failures - The State-of-the-Art. Electric Power Research Institute, Palo Alto,
Calif. EPRI CS-4329-SR, December 1985.
21. U. Bertocci. Electrochem, Metal, Vol. 3, p. 275, 1968.
22. B. C. Syrett. A Review of Corrosion of Copper Alloys Exposed to High Purity
Waters. In Proceedings, Interaction of Non-Iron Based Materials With Water and
Steam. Edited by B. Dooley and A. Bursik. TR-108236. July 1997.
23. Monitoring Cycle Water Chemistry In Fossil Plants: Volume 1, Monitoring Results.
Electric Power Research Institute, Palo Alto, Calif. EPRI GS-7556, Vol. 1. October
1991.
24. Monitoring Cycle Water Chemistry in Fossil Plants: Volume 2, International Water
Treatment Practices in Fossil Fuel Units. Electric Power Research Institute, Palo Alto,
Calif. EPRI GS-7556, Vol. 2. December 1992.
25. E. G. Brush and W. L. Pearl, Corrosion and Corrosion Product Release in Neutral
Feedwater. Corrosion, Vol. 28, No. 4, p. 129, April 1972.
26. E. G. Brush and W. L. Pearl, Corrosion Behavior of Nonferrous Alloy FeedwaterHeater Materials in Neutral Water With Low Oxygen Contents. Corrosion, Vol. 25
No. 3, p. 99, March 1969.
27. P. Burgmayer, Wheres All That Copper in Your Boiler Coming From?. Materials
Performance, Vol. 28, No. 9, p. 75, September 1989.
28. S. Sato, Study of Corrosion of Copper Alloys in High Temperature Water and
Steam - Corrosion of Copper-Nickel Base Alloys. Sumitomo Light Metal Technical
Reports, Japan, Vol. 5, p. 290, No. 4, 1964.
29. J. E. Castle, J. T. Harrison, and H. G. Masterson, The Exfoliation Corrosion of
Cupro-Nickel Feedwater Heater Tubes. British Corrosion J., Vol. 1, No. 4, p. 143,
1966.
30. C. J. Hartman and C. E. Axtell, Unusual Corrosion Problems Associated with a
Boiling Water Reactor. Presented at CORROSION/69, the Annual Conference of
National Association of Corrosion Engineers, Houston, TX, 1969.
31. O. A. Smith, A Program to Locate Copper Pickup in a High-Pressure Utility
System. Proceedings of the American Power Conference , Volume 29, p. 731, 1967.
3-23

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

32. O. I. Martynova and B. S. Rogatskin, The Behaviour of Corrosion Products in the


Feedwater Circuit of Supercritical Units. Teploenergetika, Vol. 18, No. 12, p. 65,
1971.
33. P. H. Effertz, Combined Conditioning of Water/Steam Cycle of Power Generating
Units with Continuous Flow Steam Generators Using Oxygen and Ammonia
(Combined Water Treatment CWT). Der Maschinenschaden, Vol. 53, No. 6, p. 3,
1980.
34. D. Gunn, D. McInnes, K. Knights and D. Ryan Combined Oxygen Hydrazine
Ammonia Conditioning (COHAC) and Copper Transport Through the Unit Cycle.
Interaction of Non-Iron Based Materials with Water and Steam. Edited by B. Dooley and
A. Bursik. EPRI Proceedings TR-108236. July 1997.
35. K. N. Thomsen and K. Daucik. Deposits on Turbines in Relation to the Use of
Copper Alloy Materials in the Water/Steam Cycle. Interaction of Non-Iron Based
Materials with Water and Steam. Edited by B. Dooley and A. Bursik. EPRI
Proceedings TR-108236. July 1997.
36. G. Perboni, V. Scolari and R. Zoppetti. Electrochemical Monitoring of Copper
Release From Low Pressure Pre-Heaters. Interaction of Non-Iron Based Materials with
Water and Steam. Edited by B. Dooley and A. Bursik. EPRI Proceedings TR-108236.
July 1997.
37. G. Resch and K. Zinke, Investigations Relating to the Effect of the pH-Value on the
Corrosion of Brass. VGB Kraftwerkstechnik, Vol. 60, No. 1, p. 62, 1980.
38. W. L. Pearl, S. G. Sawochka, and P. S. Wall, Corrosion Product Release from
Copper Nickel Alloys in Condenser Applications. Proceedings: 42nd Annual
Meeting, International Water Conference. Engineers Society of Western Pennsylvania,
1981.
39. N. W. Polan, G. P. Sheldon, and J. M. Popplewell, The Effect of Ammonia and
Oxygen Levels on the Corrosion Characteristics and Copper Release Rates of
Copper Base Condenser Tube Alloys Under Simulated Steam-Side Conditions.
Paper No. 81-JPGC-Pwr-9, Joint ASME/IEEE Power Generation Conference, St. Louis,
MO, Oct. 1981.
40. G. P. Sheldon and N. W. Polan, Field Testing of Power Utility Condenser Tube
Alloys. J. Mater. Energy Syst., Vol. 6, No. 4, p. 313, March 1985.
41. G. V. Vasilenko, G. P. Sutotskii, and S. A. Kozina, The Effect of the pH of the
Feedwater on Corrosion of the Low Pressure Heaters of Supercritical Power
Generating Units. Thermal Engineering, Vol. 23, p. 51, 1976.
42. E. A. Tice and C. P. Venizeios. Corrosion Behavior of Condenser - Tube Alloy
Materials. Power, Vol. 107, No. 11, November 1963.
43. W. Treitinger, Erosion Tests on Feed-Heater Tubes, Brown-Boveri Review, Vol. 54,
p. 696, 1967.
44. The ASME Handbook on Water Technology for Thermal Power System: Electric Power
Research Institute, Palo Alto, Calif. EPRI CS-6303. 1989.
3-24

EPRI Licensed Material


Corrosion of Copper Alloys in Feedwater Heaters

45. K. D. Efird Effect of Fluid Dynamics on the Corrosion of Copper-Base Alloys In


Seawater. Corrosion, Vol. 33, No. 1, January 1977.
46. B. C. Syrett Erosion Corrosion of Copper-Nickel Alloys in Seawater and Other
Aqueous Environments. A Literature Review. Corrosion, Vol. 32, No. 6, June 1976.
47. D. D. Macdonald, B. C. Syrett and S. S. Wing The Corrosion of Copper Nickel
Alloys 706 and 715 in Flowing Seawater. I. Effect of Oxygen. Corrosion, Vol. 34,
No. 9, September 1978.
48. G. Bianchi, G. Fiori, P. Longhi and F. Mazza Horse-Shoe Corrosion of Copper
Alloys in Flowing Seawater and Possibility of Cathodic Protection of Condenser
Tubes in Power Stations. Corrosion, Vol. 34, No. 11, November 1978.
49. Standards for Closed Feedwater Heaters, Fourth Edition. Heat Exchange Institute,
Cleveland, Ohio. August 1984.
50. Corrosion Related Failures in Feedwater Heaters. Electric Power Research Institute, Palo
Alto, Calif. EPRI CS-3184. July 1983.
51. V. I. Zarembo, V. G. Kritskii, A. A. Slobodov, and L. V. Puchkov, Behavior of
Copper Corrosion Products in the Water Circuits of Power Generating Units, Zh.
Prikladnoi Khimii, Vol. 62, No. 1, p. 71, 1989.
52. J. R. Meyers and A. Cohen, Pitting Corrosion of Copper in Cold Potable Water
Systems. Materials Performance, Vol. 34, No. 10, p. 60, 1995.

3-25

EPRI Licensed Material

4
ELECTROCHEMISTRY OF COPPER

INTRODUCTION
Corrosion science involves many technical disciplines, including metallurgy, chemistry,
electrochemistry, thermodynamics, kinetics, heat and mass transfer, and fluid
mechanics. The interaction between electrochemistry, thermodynamics and kinetics is
particularly important and its understanding has often created opportunities for
improved corrosion control.
All electrochemical reactions involve chemical changes and transfer of electrons
between chemical species. For example, we can consider the overall reaction of metallic
copper and oxygen in water:
2Cu + O2 Cu2O

(4-1)

At first sight, this reaction appears not to involve electron transfer. In reality, reaction
(4-1) is the sum of two other electrochemical reactions, each involving electron transfer:
one occurs at anodic (corroding) areas and the other at cathodic (non-corroding) areas
on the metal surface. The anodic reaction results in the formation of positively charged
cuprous ions (Cu+) that dissolve in the water and electrons (e-) that remain in the metal:
2Cu 2Cu+ + 2e-

(4-2)

These electrons travel through the metal to cathodic sites where they are consumed in
the cathodic reaction to produce negatively charged hydroxyl ions (OH-) that also
dissolve in the water:
O2 + H2O + 2e- 2OH-

(4-3)

The products of the anodic and cathodic reactions may then combine to yield the
cuprous oxide indicated in reaction (4-1):
2Cu+ + 2OH- Cu2O + H2O

(4-4)

4-1

EPRI Licensed Material


Electrochemistry of Copper

Anodic and cathodic areas can be in close proximity allowing Cu2O formation at the
metal surface as well as some release of Cu+ into the water.
The anodic and cathodic reactions must proceed at the same rate. Electron flow from
anodic to cathodic sites serves to maintain charge balance throughout the system.
Reduction of either the cathodic reaction (by reducing oxygen levels in the water) or the
anodic reaction (by forming a protective surface film that reduces ion flow) serves to
slow the overall reaction, thereby reducing corrosion activity.
Overall, there is a considerable body of literature in existence dealing with
electrochemistry of copper. However, much of this information focuses on
concentrated solutions including acids, brines and salt/brackish waters. Some work
has considered effects of elevated temperature but most was done under ambient
conditions. Results of such studies are most directly applicable to use of copper and its
alloys in marine environments and under chemical process conditions. Some of this
work may be useful when considering waterside corrosion of copper alloys used in
condenser tubes, or under conditions of extreme cycle contamination that leads to
concentration of impurities in deposits and crevices or at sites of excessive heat transfer.
Thermodynamic assessments determine the direction in which a reversible chemical
reaction will proceed under a set of assumed conditions. For instance, thermodynamic
analysis is used to appraise the theoretical possibility of metal dissolution or oxide
stability in a given environment. Scientists have expended considerable effort in
applying thermodynamic principles to construct stability diagrams for metals in
aqueous systems. These diagrams indicate under what conditions the equilibrium
phase of the metal is in elemental form, a metal ion, or a metal oxide or other
compound. This approach was originally implemented by Pourbaix(1); the diagrams
are usually referred to as Pourbaix diagrams or potential-pH diagrams.
POTENTIAL-pH DIAGRAMS FOR COPPER(1-14)
Figure 4-1 is a potential-pH diagram for the copper-water system at 25C, as originally
developed by Pourbaix(1) and published by the National Association of Corrosion
Engineers(2). The following species are considered as solids and soluble ions.
Solid phase: Cu, Cu2O, and CuO
-

Ions: Cu+, Cu++, HCuO2 and CuO2--

4-2

EPRI Licensed Material


Electrochemistry of Copper

Considering the solid substances Cu, Cu2O, and CuO. Cu(OH)2 is not considered.
After Pourbaix(1,2)

Figure 4-1 Potential-pH Diagram for the Copper-Water System at 25C

The diagram can be considered as three superimposed component diagrams: that for
water (lines a and b), for copper solid species in equilibrium with dissolved copper
species (solid lines), and for equilibria involving dissolved copper species alone
(broken lines). The data used to construct this diagram include the Gibbs energies of
formation of the various species involved, electrode potentials, and pH.

4-3

EPRI Licensed Material


Electrochemistry of Copper

Before interpreting this Pourbaix diagram it is useful to first briefly review the basis of
its construction. Each line represents the potential/pH relationship for a reaction at
equilibrium, which can be expressed as
aA + xH + + ze - bB + cH 2 O

(4-5)

where A and B are metal species (e.g. Cu2O and Cu, respectively), for which the Nernst
equation can be written as
Ee = Eo +

2.303RT
2.303RT x
{a log a A b log a B c log a H 2 O }
( ) pH
zF
F
z

(4-6)

where Ee is the equilibrium potential and ai is the activity (or fugacity) of species i. The
standard potential, Eo, calculated as the change in the standard Gibbs energy for the
hypothetical cell SHE/A,B, where SHE represents the Standard Hydrogen Electrode
(SHE). Thus, the cell reaction is written as
aA + (x - z)H + +

x
H 2 bB + cH 2O
2

(4-7)

and hence
G oR = b Bo + c Ho 2 O a Ao ( x z) Ho +

x o
H
2 2

(4-8)

where io is the chemical potential of species i at the temperature and pressure of


interest. The Standard Potential is then given by
E oA / B = G oR / zF

(4-9)

Equation (4-6) provides for a linear relationship between Ee and pH only if the second
term on the right side is constant. This requires that the activities of A, B, and H2O not
be a function of the independent variable, pH. The activities of A and B are set by
stipulation, which requires that the concentrations must change as
ci = ai/yi
where yi is the molar activity coefficient of the species i. From the Gibbs-Duhem
equation for multi-component systems, we may write
4-4

(4-10)

EPRI Licensed Material


Electrochemistry of Copper

N 2
pH

pH

d log a H2 O =
10
c j d log a j / cH2 O
yH +

j =1

(4-11)

pH = log a H + = log[y H + c H + ]

(4-12)

where

For dilute solutions, it is easily shown that d log a H 2 O ~ 0, so that the activity of water
remains constant, even though the pH changes over a wide range. However, in
concentrated solutions, which must exist for very high or very low pH, the activity of
water becomes a function of the ionic strength (and hence pH), so that the linear
relationship between Ee and pH breaks down. In general, this is not an issue in thermal
power plant heat transport circuits, because solute concentrations in the bulk solutions
are low. However, concentrated solutions frequently form in boiling crevices, within
sludge piles in nuclear steam generators, and even at heat transfer surfaces (in the
absence of a crevice), so that an accurate description of the thermodynamics of metals
in contact with these local (concentrated) environments may require consideration of
the non-linear behavior of the potential-pH relationship.
The interpretation of Pourbaix is straightforward, but is greatly aided by the equation
(E Ee ) I 0

(4-13)

which is a statement of the Second Law of Thermodynamics. In Equation (4-13), E is


the potential applied to the cell SHE/A,B (right hand electrode potential minus left
hand electrode potential) and I is the current. The convention is that if I>0, Reaction (45) proceeds from right to left, but if I<0 the reaction proceeds from left to right. When
I=0 (corresponding to E = Ee), the reaction is at equilibrium. Thus, with reference to
Figure 4-1, the reaction
Cu2O + 2H+ + 2e- 2Cu + H2O

(4-14)

proceeds in the oxidation sense (right-to-left), whereas


2CuO + 2H+ + 2e- Cu2O + H2O

(4-15)

occurs spontaneously in the reduction sense (left-to-right) if the potential is located


between lines 7 and 9. Equation (4-13) is quite general and can be used to predict the
direction of any half-cell reaction.
4-5

EPRI Licensed Material


Electrochemistry of Copper

A common misconception in the use of Pourbaix diagrams is that the regions of


stability of ionic species (i.e., those bounded by Lines 15, 18, and 11 for Cu+ and Lines
17, 20, and 12 for CuO2 ) correspond to corrosion, whereas the stability fields for
Cu2O and CuO imply passivity, and hence corrosion resistance. This
misconception was promulgated by Professor Pourbaix, who labeled the regions
accordingly. While under some circumstances a surface oxide may impart some
corrosion resistance, the problem is that corrosion and passivity are kinetic
concepts that have little relevance in thermodynamics except for the fact that the
reactions involved must be spontaneous.
Paradoxically, the Pourbaix diagrams are also very useful for predicting the conditions
under which metastable oxides and hydroxides form on the metal surface. Thus,
extension of Line 7 into the Cu++ stability field provides the E vs. pH conditions for the
formation of metastable Cu2O on the metal surface, as determined by Equation (4-13).
Likewise, extension of Line 9 into the Cu++ stability field yields the conditions for
formation of metastable CuO from Cu2O. Thus, if the potential lies between the
extensions of Lines 7 and 9 for a given pH in the Cu++ stability field, the copper surface
is predicted to be covered by a metastable Cu2O oxide, provided that the rates of
formation and dissolution are such that the oxide is kinetically stable. However, if the
potential is above (more positive) than the extension of Line 9, a metastable duplex film
consisting of a barrier layer of Cu2O and an outer layer of CuO or Cu(OH)2 is predicted
to exist.
In ammoniated systems, copper exhibits a dramatically different thermodynamic
behavior, as shown in Figure 4-2 for 25C. The most important change is the
appearance of stability fields for the complexed soluble species Cu ( NH 3 )2+ and
Cu(NH 3 ) 4+ + at pH values ranging from about 6.5 to 11; in the same region that, in the
absence of ammonia (Figure 4-1), Cu2O and CuO form as stable phases on the copper
surface. Furthermore, the Cu/ Cu ( NH 3 )2+ boundary extends below the extensions of the
Cu/Cu2O lines into the stability field for Cu ( NH 3 )2+ , demonstrating that conditions exist
where a soluble species may form without the possibility of forming even metastable
Cu2O (i.e., at potentials below the extensions of the Cu/Cu2O lines into the Cu ( NH 3 )2+
field from either the acidic side or the basic side of the field). In any event, ammonia
effectively depassivates copper by removal of both oxides in the pH range that is of
principal interest in feedwater systems. It is this fact that is largely responsible for the
severe corrosion problems in many steam surface condensers that employ AVT, or
which employ amines that decompose to yield ammonia. Thus, clearly, ammonia in
the feedwater is important as both a pH adjuster and a copper complexor.

4-6

EPRI Licensed Material


Electrochemistry of Copper

Source: Reference 4 as previously published in References 5 and 6.

Figure 4-2 Potential-pH Diagram for the Copper-Water-Ammonia System at 25C

High Temperature Potential-pH Diagrams for Copper


Potential-pH diagrams for copper have been constructed for elevated temperatures up
to 300C by Macdonald(15). These diagrams were constructed in the early 1970s and
employed ionic entropies calculated on the basis of the Criss-Cobble Correspondence
Principle. Since that time, thermodynamic databases have greatly improved in quality
and algorithms have been developed that construct diagrams for many metals in the
periodic table, including copper. Pourbaix diagrams for copper in aqueous
environments at elevated temperatures have been constructed at 100C, 150C, 200C,
250C, and 300C. Figures 4-3 and 4-4 show examples at 100 and 250C.
The effect of increasing temperature is to reduce the stability fields for cations (Cu+,
Cu++), increase the fields for anions ( HCuO 2 , CuO-2 ), rotate the equilibrium lines for all
reactions clockwise, and to move the Cu/Cu2O line downwards (i.e., to more negative
potentials) relative to the H2/H+ line (line a in Figure 4-1). These same trends with
increasing temperature are found for most, if not all, metals, and reflect the effect of
temperature on the quantity 2.303RT/F [see Equation (4-6)] and the decrease in the
dielectric constant of water, which results in greater stability of hydrolyzed species
4-7

EPRI Licensed Material


Electrochemistry of Copper

such as CuO2 compared with the nonhydrolyzed cations such as Cu++. The
hydrolyzed species possess much smaller surface charge densities than do the much
smaller cations of the same oxidation state, and hence have smaller electrostatic
contributions to the species chemical potentials. In the presence of complexing species,
such as Cl- and NH3, increasing temperature stabilizes the complexes [e.g., CuCl 2 and
Cu ( NH 3 )2+ ], respectively, for the same reason. However, high temperature Pourbaix

diagrams for copper in the presence of complexing species (e.g. Cl , NH 3 ) under power
plant service conditions have not been derived.

Figure 4-3 Potential-pH diagram for Cu-H2O at 100C. Activity of dissolved copper
species = 10-6 mol/kg; pressure = 1 bar.

4-8

EPRI Licensed Material


Electrochemistry of Copper

Figure 4-4 Potential-pH diagram for Cu-H2O at 250C. Activity of dissolved copper
species = 10-6 mol/kg; pressure = 1 bar

Concluding Remarks on Potential-pH Diagrams


Potential-pH diagrams provide a thermodynamic appraisal of the stability of a metal in
an aqueous solution. Ion formation (as would occur if corrosion were active, resulting
in dissolution of the metal) only appears on the diagrams in regions where such
reactions are theoretically possible. However, the kinetics of the system must also be
considered to determine if the rates of any theoretically possible corrosion reactions are
significant in a practical sense. Similarly, locations at which a stable and possibly
protective metal oxide layer is favored may not form if reaction rates are slow.
The main uses of potential-pH diagrams include predicting the spontaneous direction
of chemical reactions, estimating corrosion product composition, and predicting
positive or negative impacts on corrosion resulting from changes in the environment.
Equations and ensuing calculations made in developing such diagrams assume
equilibrium conditions. Species indicated on the diagram are those most stable for the
given potential and pH values. However, consideration of kinetics suggests that a
metal may exist in other forms, for example metastable oxides, at the stated
conditions(3). As such, the diagrams can identify regions where metals are stable in
4-9

EPRI Licensed Material


Electrochemistry of Copper

their elemental form but cannot fully define regions in which reaction of the metal with
its environment will result in an active (high rates of metal release) or passive (low
rates of metal release) state.
Finally, pure copper is not used as a structural or as a heat transfer material in
feedwater systems. Instead, because of their superior mechanical properties, various
copper-base alloys (brasses, bronzes, cupronickels) (Table 2-2) are employed.
Accordingly, it might be assumed that potential-pH diagrams would exist for the alloys
of principal interest (e.g. Cu-Zn, Cu-Ni), but this is not the case even though the
required thermodynamic data for the alloys (i.e. chemical potentials of the alloy
components) are readily available. A significant need exists for these diagrams at the
elevated temperatures of relevance to feedwater systems, because copper is frequently
alloyed with highly active metals (e.g. Zn, Al), and the alloy systems frequently deviate
strongly from ideality. These deviations negate the superposition principle, which
postulates that the potential-pH diagram for an alloy may be constructed simply as the
superposition of the diagrams for the individual alloy components, even when mixed
oxides are formed.
COPPER CORROSION KINETICS
The scientific understanding of corrosion processes is largely based on the concept of
coupled anodic and cathodic reactions as introduced by Wagner and Traud in 1938
(16). The two primary assumptions of the Wagner-Traud hypothesis (or mixed
potential theory) are as follows(14,17).
1. Any electrochemical (free corrosion) reaction consists of one or more anodic and one
or more cathodic processes (partial reactions).
2. There is no net accumulation (or loss) of charge during electrochemical reactions;
the total partial anodic current plus the total partial cathodic current is zero. Note
that there may exist an imbalance between the local partial anodic and cathodic
current densities (as for examle, in the case of pitting corrosion), but this imbalance
must be compensated for by an opposite imbalance somewhere else in the system.
For electrochemical reactions involving metals, it is further assumed that electrical
potential in the metal and in the solution adjacent to the surface at the anodic site are
equal to those at the cathodic site. In reality, there is always at least a small potential
difference, controlled in part by the electrolyte conductivity, which may in fact control
the rate of the overall reactions. Furthermore, this assumption may not be true for
situations involving electrochemical reactions with nonmetallic materials such as
semiconductors(14).
Working from this premise, it can be shown that the rates of anodic and cathodic
reactions are related to current and the ratio of anodic/cathodic surface areas. Transfer
of electrical charge occurs at the anodic and cathodic surfaces in question. This charge
transfer process may determine the reaction rate; however, mass transfer is the
4-10

EPRI Licensed Material


Electrochemistry of Copper

controlling factor in many instances. Mass transfer of reactants from the bulk solution
to the anodic and cathodic sites and mass transfer of reaction products from these sites
to the bulk solution need to be considered for all anodic and cathodic partial reactions
composing the net electrochemical reaction under consideration. The nature and
characteristics of oxide layers formed on a metal surface will influence both charge
transfer and mass transport activity(14). Corrosion processes that are rate-limited by
charge transfer are dependent on potential (as shown by the dashed line in Figure 45a). Conversely, in situations where charge transfer is rapid and the overall reaction
rate depends on mass transfer, current is not influenced by potential (as displayed by
the left and right ends of the solid line in Figure 4-5b).

Source: Reference 14

Figure 4-5 Current/Voltage Curves for Electrochemical Processes Rate-Limited by (a)


Charge Transfer and (b) Mass Transfer
Mathematical assessments included in the Wagner-Traud hypothesis demonstrate that
rates of corrosion reactions are influenced by both thermodynamics and kinetic factors.
The thermodynamic influence can be correlated to potential while the kinetics depend
upon exchange current densities. Simplified equations for corrosion potential and
corrosion current may be expressed as follows:
E corr =

[(

)(

1 e
b
E a + E ec ) 1n A a i o ,a / A c i o ,c
(
2
2

[(

)]

I corr = A a A ci o,a i o,c 1/ 2 exp E ec E ea / ( 2 b)

(4-16)

(4-17)

4-11

EPRI Licensed Material


Electrochemistry of Copper

where subscripts a and c refer to anodic and cathodic processes, respectively, Ecorr is the
corrosion potential, Ee is the equilibrium potential, io is the exchange current density,
and Aa and Ac are the areas over which the anodic and cathodic processes operate,
respectively. For these simplified equations, it is assumed that the anodic and cathodic
Tafel coefficients are equal (designated b).
The potentials of numerous oxidation-reduction reactions have been measured or
frequently calculated from thermodynamic data for temperatures ranging from 25C300C. As shown in Table 4-1(20), results are normally related to the standard hydrogen
electrode which is arbitrarily assigned a standard potential of zero (relative to itself).
Standard electrode potentials, also referred to as oxidation-reduction potentials (ORP)
or reduction-oxidation (redox) potentials, refer to electrodes in which all reactants are
at unit activity(5). For systems in which the components are not at unit activity, the
equilibrium potential (Ee) may be calculated using the Nernst equation (cf. Eq. 4-6):
E e = E o + 2.3( RT / nF) log (a ox / a red )

(4-18)

where E is the actual potential, EO is the standard potential, R is the gas constant, T is
the absolute temperature, n is the number of electrons transferred, F is the Faraday
constant, and aox and ared are the activities of the oxidized and reduced species,
respectively.
Table 4-1
Standard Electrode Potentials at 25C Versus Standard Hydrogen Electrode
Reaction
Au3+

3e-

Au =
+
+
O2 + 4H + 4e = 2H2O
Pt = Pt2+ + 2ePd = Pd++ + 2eAg = Ag+ + e2Hg = Hg2++ + 2eFe3+ + e- = Fe2+
O2 + 2H2O + 4e- = 4 OHCu = Cu2+ + 2eSn4+ + 2e- = Sn2+
2H+ + 2e- = H2
Pb = Pb2+ + 2e-

Potential, Volts

Reaction

+ 1.498
+ 1.229
+ 1.2
+ 0.987
+ 0.799
+ 0.788
+ 0.771
+ 0.401
+ 0.337
+ 0.15
0.000
- 0.126

+
Sn =
2+
Ni = Ni + 2eCo = Co2+ + 2eCd = Cd2+ + 2eFe = Fe2+ + 2eCr = Cr3+ + 3eZn = Zn2+ + 2eAl = Al3+ + 3eMg = Mg2+ + 2eNa = Na+ + eK = K+ + eSn2+

2e-

Potential, Volts
- 0.136
- 0.250
- 0.277
- 0.403
- 0.440
- 0.744
- 0.763
- 1.662
- 2.363
- 2.714
- 2.925

Source: Reference 20.

The exchange current density is defined as the rate of the reaction expressed as a
current density (A/cm2) at the equilibrium potential for the prevailing conditions. As
indicated in Table 4-2, values vary considerably(21) and must be determined
4-12

EPRI Licensed Material


Electrochemistry of Copper

experimentally. Factors influencing exchange current density include the specific


oxidation-reduction process, the electrode material, the solution chemistry, and the
system temperature.
By inspection of equations 4-16 and 4-17, it is apparent that corrosion rates are much
more dependent on current than potential. Current density data for copper in power
plant systems are unavailable and would be costly to obtain. A more practical
approach to appraising copper corrosion and release trends would appear to be
collecting data on feedwater chemistry, including copper concentrations, while closely
monitoring the system pH and potential.

Table 4-2
Exchange Current Density Values
Reaction
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
2H+ + 2e- = H2
O2 + 4H+ + 4e- = 2H2 O
O2 + 4H+ + 4e- = 2H2 O
Fe3+ + e- = Fe2+
Ni2+ + 2e- = Ni

Electrode

Solution

Al
Au
Cu
Fe
Hg
Hg
Ni
Pb
Pt
Pd
Sn
Au
Pt
Pt
Ni

2N H2 SO4
1N HCl
0.1N HCl
2N H2 SO4
1N HCl
5N HCl
1N HCl
1N HCl
1N HCl
0.6N HCl
1N HCl
0.1N NaOH
0.1N NaOH
-0.5N NiSO4

Current Density,
amp/cm2
10-10
10-6
2 x 10-7
10-6
2 x 10-12
4 x 10-11
4 x 10-6
2 x 10-13
10-3
2 x 10-4
10-8
5 x 10-13
4 x 10-13
2 x 10-3
10-6

Source: Reference 21.

Anodic Partial Reactions


The active dissolution of copper in a variety of both non-complexing and complexing
solutions has been widely studied, with all published work being restricted to ambient
or near-ambient temperatures. Extensive reviews of the subject have been published by
Bertocci and Turner(22) and by Smyrl(23), with the latter providing a valuable update to
the former.

4-13

EPRI Licensed Material


Electrochemistry of Copper

For non-complexing environments, which include sulfate and perchlorate solutions,


there is general agreement that Cu+ is the initial dissolution product formed at the
interface in a mechanism that may be represented as
Cu Cu+ + e-

(4-19a)

Cu+ Cu2+ + e-

(4-19b)

Although Reaction (4-19b) is generally regarded as being slow and hence rate
determining, there is no universal agreement on this point. Furthermore, extensive
electrochemical impedance studies of the electrodissolution of copper in sulfate media
+
suggests that adions of the type Cu ads
, which are mobile surface species, are
important intermediates in the reaction mechanism.
Because of the extensive use of copper and copper alloys in saline environments (e.g.
seawater, brackish water), it is not surprising that the dissolution of copper in the
presence of complexing species, such as chloride ion, has been extensively explored.
Studies have included determining the dissolution rate as a function of potential, pH,
flow velocity, and chloride concentration to determine kinetic parameters, such as Tafel
constants and reaction orders with respect to H+ and Cl-. Again, there is general
agreement that Cl- is an important species in the mechanistic sense, in that it strongly
couples both Cu+ and Cu2+, but there is considerable disagreement as to the details of
the reaction mechanism. For the present purposes, the dissolution of copper in acidic
chloride solutions may be represented by the following mechanism
Cu + Cl- CuClads + e-

(4-20)

CuClads + Cl- CuCl2-

(4-21)

CuClads Cu2+ + Cl- + e-

(4-22)

with Reaction (4-22) being included to account for the formation of Cu2+ at high
potentials.
The CuClads species is envisioned to form initially as a two dimensional, monomolecular film on the metal surface, particularly at low potentials, but most authors
now agree that cuprous chloride (CuCl) forms as a porous three-dimensional film. At
sufficiently low potentials the overall rate is determined by the charge transfer process,
Reaction 4-20, so that the current/voltage curve exhibits typical Tafel behavior.
However, at higher potentials the rate of the dissolution process becomes limited by the
rate of dissolution of the CuCl surface film. Since this is a chemical (not an
4-14

EPRI Licensed Material


Electrochemistry of Copper

electrochemical) reaction, the current becomes independent of applied potential. The


small residual potential dependence may be attributed to the production of Cu 2+
(Reaction 4-22).
Because of the ubiquitous presence of water, the possibility that OH- can act as a
complexing species must be recognized and indeed the experimental reaction order
data with respect to H+ (and hence OH-) at relatively high pH (pH = 3 to 8) supports
this hypothesis. Indeed, the data are consistent with the gradual development, with
increasing pH, of the mechanism
Cu + H2O CuOHads + H+ + e-

(4-23)

CuOHads + H+ Cu+ + H2O

(4-24)

Cu+ + 2Cl- CuCl2-

(4-25)

Thus, Cl- and OH- may be viewed as competitive species in the initial electrochemical
adsorption step, with Cl- adsorption dominating at high [Cl-] and low pH but with OHadsorption occurring at low [Cl-] and low pH.
Not all species that complex Cu+ (or Cu2+) necessarily are involved in initial
electrochemical adsorption steps in the dissolution of copper. Perhaps the most
important species in this class is ammonia (NH3), which is involved in the active
corrosion of copper alloys in condensers and feedwater heaters in both fossil and
nuclear power plants that operate on AVT. Although (and somewhat surprisingly) the
kinetics of copper and copper alloy dissolution in ammoniated environments that are
relevant to thermal power plants do not appear to have been studied in sufficient depth
to determine reaction orders with respect to the various species in the system (e.g. NH3,
H+), it is likely that ammonia acts simply as a complexing species in a mechanism of the
type:
Cu Cu+ + e-

(4-26)

Cu+ + 2NH3 Cu( NH 3 )

(4-27)

Cu(NH3)+ Cu( NH 3 )

(4-28)

+
2

2+
+ e
2

to yield the overall partial anodic reactions:


4-15

EPRI Licensed Material


Electrochemistry of Copper

Cu + 2NH3 Cu( NH 3 ) + 2e

(4-29)

Cu + 2NH3 Cu( NH 3 )

(4-30)

+
2

and
2+
+ 2e
2

which occur at low ORP and high ORP, respectively.


Elsewhere in this report, it is noted that the corrosion of copper alloys is particularly
severe in ammonium carbonate solutions, which form via air in-leakage in cycles that
operate on AVT. Indeed, the corrosion rate appears to be higher than can be accounted
for by the presence of NH3 and/or NH4+ alone, suggesting that carbonate ion may play
a catalytic role in the reaction mechanism. There does not appear to be any reliable
data that indicate the formation of carbonate complexes of the type Cu(CO3)- or
Cu(HCO3)0, which would seem to leave electrochemical adsorption (c.f. Cl-) as the only
feasible mechanistic involvement. On the other hand, CO2 in-leakage may simply act to
modify the speciation (e.g. NH3/NH4+) in the bulk environment through changes in pH
and not to have any specific electrochemical effect on the dissolution kinetics of the
copper alloy. This issue needs to be resolved, because if carbonate species are involved
electrochemically in the partial anodic reaction the problem of enhanced copper alloy
corrosion due to air in-leakage will probably only be solved by maintaining tighter
condensers. However, if CO2 serves to modify speciation, then the problem may well
be countered by modifying the AVT.

Cathodic Partial Reactions


As noted, corrosion processes consist of one or more partial anodic reaction, and one or
more partial cathodic reaction, and many examples exist where the partial cathodic
reaction(s) control the rate of the overall (corrosion) reaction. The cathodic reactions of
principal interest are the reduction of O2, H2O2 (which may be formed as an
intermediate in the reduction of oxygen), and Cu( NH 3 )

2+
[which forms in solution via
2

the homogeneous oxidation of Cu( NH 3 ) . The reduction of H+ (i.e. H + + e 1 / 2 H 2 )

+
2

and H 2O H 2 O + e OH +1/ 2H 2 is commonly thought to be unimportant in copper


corrosion, because of the semi-noble nature of the metal. However, in the presence of
strong complexing agents or activating species, such as NH3 and HS-, respectively, the
Cu/Cu(I) equilibrium line may be shifted in the negative direction to the extent that
hydrogen evolution from either the reduction of H+ or the reduction of water becomes a

4-16

EPRI Licensed Material


Electrochemistry of Copper

viable cathodic reaction. In this case, copper will corrode in anoxic environments and
hence the metal loses its semi-noble character.
Hydrogen evolution on copper has been studied extensively and, in keeping with the
semi-noble character of copper, the reaction is commonly assumed to occur via the
same mechanism that occurs on the noble metals, such as platinum, viz:
H+ + e- Hads

(4-31)

Hads + Hads H2

(4-32)

However, consensus with regards to the latter step does not exist, and various authors
insist that the desorption reaction occurs as the irreversible process
H+ + Hads + e- H2

(4-33)

Still others insist that both Reactions (4-32) and (4-33) occur simultaneously at rates that
are comparable. Thus, the mechanistic situation with respect to the hydrogen electrode
reaction on copper (and more so on copper alloys) is very confused, with a lack of
consensus as to even the nature of the elementary reactions that occur at the surface.
The kinetics of the oxygen reduction reaction on copper has also been explored with the
most significant work being reported by Vazquez et al(24). These workers used rotating
disk voltammetry coupled with Koutecky-Levich analysis to obtain kinetic parameters
(Tafel constant, reaction orders), which were found to be consistent with the
mechanism

(4-34)

with k3 >> k4, being required to account for the observed small yield of H2O2 in the
solution. However, not all workers who have studied this reaction report that the yield
of hydrogen peroxide is small(23). Part (if not all) of the difficulty may lie in the fact that
the homogeneous decomposition of hydrogen peroxide is catalyzed by heavy metal
ions, so that the overall mechanism may be autocatalytic, as indicated by the broken
line path in the above mechanism (our addition).

4-17

EPRI Licensed Material


Electrochemistry of Copper

Vazquez et al(24) have also explored the kinetics and mechanisms of hydrogen peroxide
reduction on copper in aqueous solution. The kinetics of this reaction were found to be
sensitive to the extent of prior oxidation of the copper surface, and unusually large
Tafel constants were found for oxidized surfaces. We suggest that the latter
observation simply reflects that part of the applied potential occurs as a potential drop
across the passive film and hence that only part of the applied voltage is available for
driving the H2O2 reduction reaction at the film/solution interface. According to
Vazquez et al(24), the reduction of H2O2 occurs via a redox cycle that can be represented
by the following two reactions:
Cu2O + H2O2 2CuO + H2O

(4-35)

2CuO + H2O + 2e- Cu2O + 2OH-

(4-36)

to yield the overall reaction


H2O2 + 2e- 2OH-

(4-37)

The catalytic involvement of Cu2O in the reduction of H2O2 was postulated principally
on the observed effect of chloride ion on the kinetic current; chloride was observed to
decrease the current over most of the potential range. The authors(24) attribute the effect
of chloride ion to the destruction of the Cu2O catalytic phase via the reduction
Cu2O + 4Cl- + H2O 2CuCl2 + 2OH-

(4-38)

The final reaction that needs to be discussed is the reduction of the cupric ammonium
couplex ion, Cu( NH 3 )

2+
, which has been implicated in the corrosion of copper-base
4

alloys in condensers in units operating under AVT. This reaction is best represented as
occurring via a catalytic chemical-electrochemical (EC) mechanism
(4-39)

where the broken line represents the path for the catalytic generation of Cu( NH 3 ) .
+
2

Addition of the above reactions yields

4-18

EPRI Licensed Material


Electrochemistry of Copper

1
1
O2 + H + + e H 2O
4
2

(4-40)

which is the normal oxygen reduction reaction. Thus, the Cu( NH 3 ) / Cu( NH 3 )
+
2

2+
redox
4

couple acts as a couple. To the authors knowledge, this important reaction has never
been subjected to careful kinetic and mechanistic analysis. Thus, it is not actually
known that the reaction can be represented by Scheme (4-39) and no kinetic data are
currently available.
In summary, the state of knowledge of the cathodic reactions that occur on copper and
copper alloys that are of interest to the thermal power industry is exceedingly poor.
Thus, the mechanisms of the reactions are poorly-defined and no kinetic data of the
type that are required for modeling purposes are available. This is true for copper
alloys under conditions that are relevant to thermal power plant steam cycles. No
quantitative corrosion modeling, and hence life prediction, is possible unless the kinetic
data are measured.
PASSIVITY OF COPPER
The relationship between applied potential and resulting current may be linear over
narrow ranges, particularly within about 10mV of the open circuit potential (i.e. the
corrosion potential). However, at potentials further from the open circuit potential, the
potential is proportional to log (current) for reactions limited by charge transfer. The
introduction of surface films and mass transfer limitations creates other relationships
between potential and current.
Decreasing the potential below the open circuit potential normally stimulates oxygen
reduction or hydrogen evolution reactions at the metal surface. Increasing the
potential, on the other hand, normally results in an increase in the metal dissolution
rate (the dissolution region in Figure 4-6). However, it is frequently observed that
with further increases in potential above some critical value (the passivation potential,
Ep in the figure) a film forms over the metal surface and current diminishes. The
potential at which current is minimized is referred to as the Flade potential (EF). The
passive film is generally protective at voltages somewhat higher than the Flade
potential. However, further increases eventually result in the system entering a
transpassive state and, ultimately, a condition under which oxygen evolution occurs.

4-19

EPRI Licensed Material


Electrochemistry of Copper

Source: Reference 14

Figure 4-6 Schematic Current/Voltage Curve for a Passive Metal

Field Observations
Until recently, and with the exception of work by Macdonald to evaluate the
electrochemical behavior of copper in lithium hydroxide solutions(25), minimal work
has been done to assess passivity and passivity breakdown for copper under feedwater
heater temperatures. Development studies designed to improve scientific
understanding of the electrochemistry of copper in the feedwater heater environment is
considered an area deserving of additional research effort(3). Extrapolation of existing
low-temperature data may not be satisfactory due to temperature effects on solution
chemistry, solubility of copper and its oxides, and reaction kinetics.
Passivity assessments for copper at low temperatures have been more extensive.
Generally, solution chemistries have been selected to assess effects of constituents
which may be present in natural waters (such as once-through cooling water). Overall,
these investigations have tended to verify the nature of copper films as predicted by
thermodynamic analysis. As discussed in Section 3, in aqueous solutions which are
free of sulfide ion, a copper film consisting of two distinct layers is typically formed.
The inner layer, adjacent to the metal, consists of cuprous oxide (Cu2O). The
composition of the outer layer is dependent on solution chemistry and may include
cupric oxide (CuO), cupric hydroxide [Cu(OH)2], and chloride, sulfate and carbonate
species(3). Anions in the aqueous solution influence the passivity of copper films as well
as the solubility of copper ions and compounds.
An investigation of failures of admiralty brass tubes in low pressure feedwater heater
service at Ontario Hydro determined that localized loss of metal at the tube bends were
being experienced due to an erosion-corrosion (or flow-accelerated corrosion)
mechanism(26). Some important findings and conclusions of this work are as follows:

Metallurgical assessments verified that damage and failure involved a flow-assisted


localized corrosion mechanism which was electrochemical in nature.

4-20

EPRI Licensed Material


Electrochemistry of Copper

Laboratory testing determined that an electrochemical cell existed between bare


metal (corroded) surfaces and passive areas (covered by oxide scale and
deposits).

Activity in the resultant cell was increased by elevations in water temperature and
concentration of dissolved oxygen, ammonia and carbon dioxide.

Morpholine (used for pH control) additions produced increases in copper release


rates over the short term, proportional to the injection rate; substitution of ammonia
for morpholine did not cause an increase in copper release.

Carbon dioxide additions resulted in significant, instantaneous releases of copper.

Increased copper release while using morpholine was attributed to species found as
a result of thermal decomposition including ammonia, organics and carbon dioxide.

Laboratory experiments to measure current densities under controlled conditions


verified significant differences for bare metal and surfaces covered with oxide scale
or deposits and demonstrated the effects of feedwater chemistry parameters on
corrosion activity.

Upon completion of the assessment, a recommendation was made that all stations
replace morpholine with ammonia for feedwater pH control. This recommendation
was accepted and has been followed since 1985.
Film formation, passivity and passivity breakdown are also influenced by dissolved
gases in the water. Research performed in Italy determined that the protectiveness of
surface films found on copper in flowing high purity water at 40C depends on the
partial pressures of hydrogen, oxygen and carbon dioxide and is further influenced by
the specific exposure history(27). Other investigations(2831) have concluded that the
nature and stability of the film varies with the oxygen concentration. Some important
conclusions of these studies are as follows.

Cuprous oxide (Cu2O) formed in flowing water with low oxygen levels (<30 ppb)
tends to be loosely adherent and is swept along by the water flow as a colloidal
suspension; any deposits formed on downstream surfaces are unprotective(28).

At higher oxygen levels (>100 ppb), an adherent layer of dark cuprous oxide (Cu2O)
will form; this layer provides some protection of underlying metal(28). However,
continued exposure leads to formation of an outer layer of cupric oxide (CuO) or
cupric hydroxide [Cu(OH)2] which is only loosely adherent(29).

At oxygen concentrations of 2-3 ppm, a relatively protective surface layer consisting


principally of cupric oxide (CuO) is formed(28,30); any cuprous oxide formed initially
is rapidly converted to cupric oxide(31), although it is most likely that a thin Cu2O
barrier layer continues to exist in contact with the metal.

Release rates for particulate copper oxides tend to fall with increased oxygen levels,
however, release rates for dissolved copper increase(30).
4-21

EPRI Licensed Material


Electrochemistry of Copper

In essence, work on oxygen shows the relationship between oxygen concentration,


passivity and corrosion/release of copper to be complex in nature. The dynamic
conditions which exist in an operating fossil unit feedwater heater train further
complicate the situation and effectively minimize the direct applicability of the research
findings. Furthermore, the copper alloys actually used in feedwater systems behave
differently from elemental copper.

Mechanisms of Passivation and Passivity Breakdown


Our understanding of the fundamental processes that lead to passivity and passivity
breakdown on copper in aqueous environments is relatively good, particularly since
the development of the Point Defect Model(32), with the original work being supported
by EPRI (RP1166-1). Thus, extensive X-ray Photoelectron Spectroscopy (XPS) studies
have shown that the passive film on copper consists of a Cu2O barrier layer adjacent to
the metal and an CuO or Cu(OH)2 outer layer that forms by precipitation of cations that
flow through the barrier layer or are generated by dissolution at the barrier layer/outer
layer interface (Figure 4-7). The Point Defect Model(32) (PDM) seeks to describe the
reactions that occur in the system in terms of the generation or annihilation of cuprous
'

ion vacancies VCu


and oxygen vacancies VO
at the metal/barrier layer and barrier
layer/outer layer interfaces, as depicted in Figure 4-8. Note that the transmission of
cuprous ions through the Cu2O barrier layer is described by Reactions (1) and (3),
which when added together yield
Cu Cu 2 + + VCu + 2 e

(4-41)

corresponding to dissolution of the metal and to the formation of a metal vacancy at the
Cu/Cu2O interface. On the other hand, the addition of Reactions (2) and (4) to
eliminate the oxygen vacancy yields
2Cu + H 2 O Cu 2 O + 2 H + + 2 e

(4-42)

which corresponds to the formation of the barrier layer. Thus, it can be concluded that
'
the movement of cuprous cation vacancies ( VCu
) across the film leads to metal

dissolution whereas the movement of oxygen vacancies ( VO


) leads to barrier layer
growth.

4-22

EPRI Licensed Material


Electrochemistry of Copper

Figure 4-7 Schematic of processes that lead to the formation of bilayer passive films on
metal surfaces. Source: Reference 32.

Figure 4-8 Schematic of processes that lead to the formation of the Cu2O layer and
the Cu(OH)2/CuO outer layer in the passive film on copper in oxidizing aqueous
'

media. VCu
= cuprous ion vacancy in Cu2O, CuCu = cuprous cation in Cu2O, VO
=
oxygen vacancy in Cu2O. This scheme assumes that the barrier layer is a cation
vacancy conductor; a comparable scheme can be constructed for an interstitialconducting barrier layer.

4-23

EPRI Licensed Material


Electrochemistry of Copper

An important property of the reactions that occur at the interfaces shown in Figure 4-8
is whether or not they are lattice conservative. Careful analysis shows that Reactions
(1), (3), and (4) are lattice conservative, in that their occurrence does not result in any
movement of the corresponding interface with respect to a fixed reference frame
outside of the system. On the other hand, Reactions (2) and (5) are lattice non-

conservative, in that the former describes the formation of Cu 2 O i. e. Cu 0.5 V0 from


copper while the latter accounts for the oxidative dissolution of the barrier layer to
form Cu2+ in solution or CuO/Cu(OH)2 in the outer layer. Experiments show that the
barrier layer exists in a steady-state, in which case the rates of Reactions (2) and (5),
Figure 4-8, must be equal. This condition leads to the following expressions for the
steady-state barrier layer thickness and current of
k0
L ss = aV + bpH + c ln 20
k5

(4-43)

and

I ss = 2 F k 03 e a 3 v e b 3 pH + k 50 e a d v e b d pH C nH +

(4-44)

respectively, where a, b, c, a3, b3, ad, and bd are constants that are determined by
fundamental constants of the system, and n is the kinetic order of the barrier layer
dissolution reaction with respect to the hydrogen ion concentration at the barrier
layer/outer layer interface (note that the outer layer is considered to be a porous
phase). Note that the steady state current (Iss) is determined by two terms: the
transmission of cuprous ions across the barrier layer and their subsequent ejection at
the barrier layer/outer layer interface (first term) and dissolution of the barrier layer
(second term), which must be balanced by the formation of new barrier layer at the
Cu/Cu2O interface via Reaction (2), Figure 4-8. Finally, note that the thickness of the
barrier layer is predicted to vary linearly with the applied voltage (or ORP), which is
observed experimentally, as well as being determined by pH and the ratio of the
standard rate constants for film formation (k20) and film dissolution (k50).
The reader will note that the steady-state current is predicted to be the sum of two
exponential functions of voltage. However, the barrier layer on copper is found to be a
p-type semiconductor, indicating that the principal point defect is the cuprous ion
'
vacancy, VCu
. Thus, the first term in Equation (4-44) dominates, so that the steadystate current is predicted to exhibit a Tafel relationship with respect to voltage

lnI ss = ln 2 Fk 03 + a 3 V + b 3 pH
4-24

(4-44)

EPRI Licensed Material


Electrochemistry of Copper

Numerical analysis shows that a3 is very small (~3.9 in the case of Ni), so that Iss is only
weakly dependent on V (corresponding to an increase in the current by about a factor
of ten on increasing the voltage by 0.7V). This value may be compared with a typical
value of a3 > 20 for the active dissolution of a metal under Tafel conditions. Although
the electrochemical properties of the barrier layer on copper (in the absence of the outer
layer) apparently have not been studied, the voltage dependence of the current in the
oxide formation region is consistent with the PDM, as outlined above.
Now returning to consider the copper release rate. If the entire interphase (i.e. the
barrier layer and outer layer) exist in a steady state then the rate of copper release to the
environment must be given by the steady-state current, in which case
dmCu
= k 03 ea 3 V e b 3 pH
dt

(4-45)

assuming that cation (Cu+) transmission dominates over barrier layer dissolution (i.e.
that the barrier layer is a p-type semiconductor), where mCu is the number of moles of
copper released per square centimeter of surface. Equation (4-45) predicts that the
copper release rate will increase exponentially with voltage (ORP), which is consistent
with the observed increase in copper release rate with increasing oxygen
concentration(27). The value of b3 could be positive or negative, depending upon the pH
relative to the pH of zero charge of the surface. At the present time, too few data of a
fundamental nature exist to permit an a priori estimation of b3 and hence to predict how
the release rate might be affected by pH. Nevertheless, the above theory is believed to
capture the principal constitutive relationships between release rate and system
properties (V, pH), with the exception of flow velocity. It is believed that flow rate
exerts its principal affect on the outer layer, which consists of poorly adherent
precipitated material [CuO/Cu(OH)2]. Thus, increasing the flow rate will tend to
shear-strip the outer layer from the surface, thereby leading to enhanced levels of
particulate copper-containing material in the feedwater. However, increasing flow rate
also has an important indirect effect in its impact on the corrosion potential. Thus,
noting that the rate of corrosion of copper in oxygenated environments is dominated by
the transport of oxygen to the surface, the Wagner-Traud hypothesis predicts that
increasing flow rate will cause the corrosion potential to shift in the positive direction; a
prediction that has been confirmed for copper-nickel alloys in oxygenated seawater(33).
According to Equation (4-45), increasing corrosion potential implies an increasing
copper release rate.
The Point Defect Model(32) has also been extensively developed to account for passivity
breakdown phenomena and a brief discussion of this subject as it pertains to the
corrosion of copper and copper alloys is given below. The central hypothesis is that
passivity breakdown initiates by vacancy condensation at the metal/barrier layer
interface in a sequence of events that is captured by the cartoon shown in Figure 4-9.
Returning now to Figure 4-8, it is noted that cation vacancies are generated at the
4-25

EPRI Licensed Material


Electrochemistry of Copper

barrier layer/outer layer interface by Reaction (3), but are annihilated at the
metal/barrier layer interface by Reaction (1). If the rate of generation of vacanacies by
Reaction (3) is increased, for whatever reason, and the rate is such that the vacancies
arriving at the metal/barrier layer interface cannot be annihilated by Reaction (1), the
excess vacancies will condense and cause local detachment of the film from the metal.
Consequently, Reaction (2) cannot occur, so that the barrier layer does not continue to
grow into the metal; nevertheless the barrier layer continues to dissolve at the barrier
layer/outer layer. Therefore, the barrier layer thins locally and eventually ruptures to
mark an initiation event. Most such events are believed to immediately repassivate,
thereby giving rise to current noise that is attributed to metastable pitting. However
in rare cases, the pit embryo is able to achieve critical conditions that allows it to
develop into a stable pit.

Figure 4-9 Cartoon outlining various stages of pit nucleation according to the PDM.
Source: Reference 32.
The processes that are postulated to lead to the enhanced ejection of cations from the
barrier layer are depicted in Figure 4-10. In all cases, the initial event is believed to be
the absorption of an aggressive species (e.g. Cl-, but also CO 2
3 ) into the oxygen
vacancy structure at the barrier layer/outer layer interface, followed by either a MottSchottky pair reaction to generate additional oxygen vacancies and cation vacancies
(Route I) or by cation vacancy submergence through the outward movement of Cu+
through the barrier layer with the result that oxygen vacancies are regenerated at the
interface (Route II). In both cases, the regeneration of oxygen vacancies renders the
processes autocatalytic and hence not subject to saturation. Furthermore, oxygen
4-26

EPRI Licensed Material


Electrochemistry of Copper

vacancy generation, and hence absorption of X-, is predicted to occur above the point of
cation vacancy condensation, in this case at the periphery of the developing condensate
(blister). This is precisely the observation of Bargeron and Givens(34), who found
chloride ion concentrated on the outer surface of the barrier layer at the periphery of
the developing blister that eventually ruptured to form a pit.

Figure 4-10 Summary of proposed reactions to passivity breakdown.


Source: Reference 32.
The above sequence of events [specifically Route (I)] leads to a prediction of the critical
breakdown voltage (Vc) and induction time (tind) for V > Vc in the form of
Vc = Vc0

( )

2.303RT
log a x F

(4-46)

and
XFV
t ind = exp
1
2 RT
'

(4-47)

where VC0 is the standard breakdown voltage (for a x = 1 ), is the dependence of the
voltage drop across the barrier layer/outer layer interface on the applied voltage, X is
4-27

EPRI Licensed Material


Electrochemistry of Copper

the oxide stoichiometry (X = 0.5 for Cu2O), is the time required to thin (by
dissolution) the barrier layer cap above the vacancy condensate to the point where
rupture occurs, and V = V - Vc is the applied overpotential. This theory has been
further developed(32) to predict distributions in Vc and tind by assuming a normal
distribution of the breakdown sites with respect to cation vacancy diffusivity within the
barrier layer. These distribution functions, in turn, have been combined with pit
growth models to predict, deterministically for the first time, damage functions for
pitting corrosion. However, since this type of analysis has not been applied to copper,
no further discussion is given here, but the interested reader could refer to the literature
for additional material(32).
The most extensive experimental investigation of the pitting of copper under conditions
that are relevant to the thermal power industry is probably that of Milosev et al.(35)
These workers studied the pitting of copper in HCO 3 / SO 24 solutions by measuring
both the breakdown voltage and the induction time over wide ranges of compositional
variables. They interpreted their data in terms of several current theories for passivity
breakdown, but found that only the PDM gave a satisfactory account of their

experimental findings. Their analysis yielded = 0.352-0.374, = 5.16s, and VC0 = 1.82 V vs SCE (for Na2SO4 solutions without bicarbonate). Substitution of these values
into Equations (4-46) and (4-47), and noting that X SO 24 , enables one to estimate Vc
and tind for the pitting of copper sulfate in bicarbonate solutions. However, it is
important to note that calculated values of Vc and tind should be identified with the
means (assuming a sufficiently large number of experiments was performed), because
the distributed nature of the breakdown sites was not recognized in the study. Finally,
as in the earlier work of Strehblow and Titze(36), HCO 3 was found to inhibit the pitting
of copper by a process that is not yet fully understood.
The reader will recall from an earlier section of this report that studies by Macdonald,
Syrett, and Wing(33) showed that in the corrosion of copper-nickel alloys in oxygenated
seawater, increasing flow rate shifted the corrosion potential in the positive direction
and that the critical flow velocity gave a corrosion potential that coincided with the
voltage at which transpassive breakdown occurs in quiescent systems. This
observation led to the suggestion that the reason for flow-induced breakdown is
fundamentally electrochemical in nature and not mechanical. In other words, it is best
described as being a flow mediated electrochemical phenomenon. Because of the
importance of erosion-corrosion in copper-nickel alloy performance in thermal power
plants, a brief discussion of this topic in terms of the theory of passivity breakdown is
warranted.
Returning to Figure 4-8, it is noted that Reaction (3) which generates cation vacancies
'
VCu
at the barrier layer/outer layer interface is electrochemical in nature (i.e. it
4-28

EPRI Licensed Material


Electrochemistry of Copper

involves the production of electrons). Indeed, the rate constant for this anodic reaction
can be written in the form
k 3 = k 30 e a 3 V

(4-48)

where a3 > 0. Thus, as the voltage increases, due to increasing flow velocity, the rate of
production of cation vacancies at the barrier layer/outer layer interface also increases.
At some critical voltage, corresponding to a critical flow velocity and hence a critical
potential, even in the absence of an aggressive species, the enhanced flux of cation
vacancies across the barrier layer cannot be accomodated by Reaction (1) and vacancy
condensation occurs. Because the absorption of an aggressive species at specific sites
(the so-called weak spots) at the barrier layer/outer layer interface is not involved in
this breakdown process, breakdown is not highly localized, as in pitting corrosion, but
instead tends to occur over large areas corresponding to erosion-corrosion.
At this point, it is of interest of explore, on the basis of the PDM, how a copper alloy
may be made more resistant to passivity breakdown and less prone to the release of
copper species into the environment. The question of enhanced passivity has been
addressed by Urquidi and Macdonald(37) in the form of the Solute-Vacancy Interaction
Model (SVIM). Briefly, this model, which is contained within the PDM, notes that any
factor that decreases the diffusivity and/or concentration of cation vacancies within the
barrier layer will result in a shift in Vc in the positive direction and an increase in tind.
The SVIM proposes that this can be done by alloying. Thus, consider an alloying
element having an oxidation state that is greater than that of the host cation [e.g. Cr(VI)
in Cu2O], and assume that the element segregates substitutionally in the oxidized form
5
into the barrier layer. The species that is present can then be represented as CrCu
,
6+
which in Kroger-Vink notation designates a Cr ion substituted into a cuprous cation
vacancy. The immobile substitute species can then interact electrostatically with mobile
'
5
VCu
, thereby decreasing the
cation vacancies to form complexes of the type CrCu
diffusivity and concentration of free vacancies and hence rendering the system more
resistant to breakdown. Indeed, a first-principles calculation of the effect of
molybdenum on the pitting potential of 18Cr8Ni (i.e. Type 304-316 SS) rendered
predictions that are in good agreement with experiment(37). The same theory appears to
account for the remarkable effects observed on the addition of small amounts of
chromium to copper-nickel alloys on their resistance to erosion-corrosion (Figure 4-11),
although no detailed modeling of the Cu-Ni-Cr system has been carried out. The effect
of small additions of Cr to Cu-Ni alloys on their resistances to flow-induced passivity
breakdown is most important, because the additions are small enough that there should
be minimal impact on the thermal conductivity of the alloy (and hence minimal impact
on the required heat transfer area for a given power output of a plant), but yet are
sufficiently large to substantially enhance the resistance of the alloy to passivity

4-29

EPRI Licensed Material


Electrochemistry of Copper

breakdown and copper release. Indeed, alloys of this type should be considered prime
candidates for condenser and feedwater heater retubing.

Figure 4-11 Effect of Chromium Additions on Seawater ImpingementCorrosion


Resistance of Copper-Nickel Alloys. 36-Day Test with 7.5 m/s Jet Velocity; Seawater
Temperature: 27C. Source: Reference 14
We should note that the PDM predicts yet another way that the resistances of metals
and alloys to passivity breakdown can be enhanced; by photoquenching of the electric
field in the barrier layer. This is predicted to shift Vc to more positive potentials and tind
to longer times. Under EPRI sponsorship (RP8041-07), this phenomenon was explored
for a wide range of substrates (Fe, Ni, Cu, stainless steels, Cu-Ni alloys) under UVenhanced tungsten lamp irradiation and under laser ( = 325 nm) irradiation. The
effect occurs only for photons having energies that are greater than the bandgap of the
oxide comprising the barrier layer, which is consistent with the primary event being the
generation of electron-hole pairs. Iron was found to exhibit extra-ordinarily large shifts
in Vc (>800 mV) upon irradiation with laser light at power densities up to 1680
mW/cm2, and indeed it is possible to render iron stainless by this treatment.
Smaller, but nevertheless significant shifts in Vc were observed for the other metals and
alloys, with that for 90Cu10Ni (CDA-706) being 40-60 mV under tungsten lamp
irradiation. A remarkable feature is that the photoinhibition effect persists for more
than 200 hours after irradiation is stopped. This is attributed to an initial
photoquenching of the electric field (by electron-hole pair production), which in turn
modifies the vacancy structure. Because the vacancy structure relaxation time is
determined by the vacancy diffusivity (D), which is of the order of 10-19 cm2/s, the
recovery is very slow, taking about 250 hours (as estimated by t ~ L2/D, where L = 30 x
10-8 cm is the thickness of the barrier layer). Although a shift in Vc of 40-60 mV may
seem to be quite modest, it does result in a lengthening of the induction time for
passivity breakdown by up to an order in magnitude. For corrosion processes (e.g.
4-30

EPRI Licensed Material


Electrochemistry of Copper

pitting) that are frequently initiation-dominated, the possible benefits of


photoinhibition are therefore clear.
It is evident from the above that the theory of passivation and passivity breakdown is
well-developed, with EPRI having provided significant support at critical stages in the
development of the subject. The theory (PDM/SVIM) has been extensively tested and
the authors are not aware of any instance where the predictions have not been
confirmed by experiment. With regards to copper alloys, the principal problem is that
the theory has never been used to address corrosion issues that are relevant to copper
alloys that are employed in the thermal power industry. The little quantitative work
that has been reported (e.g. that by Milosev et al(35)) has been carried out on pure
copper. The other examples are of an anecdotal nature or were performed to
demonstrate an effect (e.g. photoinhibition) rather than for formulating corrosion
protection strategies.
CORROSION MONITORING BY ELECTROCHEMICAL TECHNIQUES
Researchers have shown that dissolution of metals naturally induces small changes in
the corrosion potential and corrosion current, measurable as potential noise and
current noise, respectively(3840). The noise level and characteristics have been
correlated with the rate and mechanism of corrosion. Ultimately, it may be feasible to
apply the measurement methodologies to monitor on-line copper corrosion/passivity
breakdown in fossil plants(3). Successful development of such a device, supplemented
by potential-pH measurements, and conventional power plant chemistry analyses,
would provide a much improved basis for control of copper alloy corrosion activity.
Part II of the Reference Manual for On-Line Monitoring of Water Chemistry and
Corrosion provides an excellent discussion of available and emerging techniques for
surveillance of corrosion and potential(38). Physical and electrochemical measurement
techniques presented in the manual are listed in Table 4-3.
Commercially available on-line corrosion monitors which employ these techniques
usually are intended for use under low temperature service conditions. Most utility
applications are currently in cooling and service water systems(38).
The manual indicates that there is limited utility industry experience with on-line
monitoring of corrosion potentials and oxidation-reduction potentials in high
temperature aqueous environments. All applications were in foreign light water
reactor feedwater systems. The monitors used were manufactured in Japan and
Europe(38).
Table 4-3
Available Techniques for Monitoring Corrosion and Potential

4-31

EPRI Licensed Material


Electrochemistry of Copper

Physical Methods

Electrical Resistance

Electrical Field Mapping

Metal Activation

Electrochemical Methods

Linear Polarization Resistance

Electrochemical Impedance Spectroscopy

Zero Resistance Ammetry (Coupling Current)

Electrochemical Potential Noise

Electrochemical Current Noise

Electrochemical Resistance Noise

A section in the manual on emerging technologies indicates that materials and


temperature measurement techniques are now available to permit manufacture of high
temperature corrosion monitors employing the electrical resistance method(39).
Electrochemical methods including electrochemical noise, electrochemical impedance
spectroscopy and zero resistance ammetry were applied over a seven month period to
monitor fireside corrosion in the superheater of a coal-fired boiler(39,40).
New developments in electrochemical monitoring and its application to power plants
were discussed during an Electric Power Industry Workshop held in conjunction with
the 12th International Corrosion Conference, sponsored by NACE International(41).
Topics covered included electrochemical monitoring of erosion-corrosion, use of
electrochemical noise methods, electrochemical sensors and monitoring for nuclear
power plants, and development of high temperature sensors for in-situ monitoring.
Particularly interesting was the paper which described work done at the St. Lucie
Nuclear Plant, describing the effect of hydrazine and dissolved oxygen levels on
electrochemical potential readings and iron release and transport(42).

ENEL Experience
In consideration of generally favorable experience worldwide with oxygenated
treatment (OT), ENEL of Italy implemented the treatment at a number of fossil power
stations. Some of the units included copper alloys in the low pressure feedwater
heaters. Extensive testing for copper was conducted at the Piombino Station in the
early 1990s. Experiences in one unit which was converted to OT from all volatile
treatment (AVT) and a second unit which was started up on OT were described at an
international conference in 1996(43). During the evaluation, extensive measurements of
dissolved and particulate copper were made at several points in the cycle. Copper
4-32

EPRI Licensed Material


Electrochemistry of Copper

transport rates were highly dependent on test conditions. (Copper transport is the
subject of the next section of this report.)
More recently, ENEL initiated an electrochemical monitoring program to evaluate
release of copper from low pressure heaters under various chemistry conditions(44).
Treatments evaluated included AVT and OT, utilizing hydrogen peroxide and oxygen
as the oxidants. Copper solubility in feedwater was predicted with a thermodynamic
model called EQUAT(45) and compared to analytical data for field samples.
The unit on which monitoring was conducted included three low pressure heaters with
tubes of arsenical copper. High temperature electrochemical sensors were installed at
the outlet of the third heater which typically cycles between 100-120C. Copper
electrochemical potential, corrosion rate (via linear polarization resistance
measurements) and specific conductivity were measured at feedwater temperature.
The equipment used was obtained from a foreign supplier and is based on technology
developed by EPRI. In addition, the cooled sample was monitored for dissolved
oxygen, hydrazine, and specific and cation conductivities.
Important conclusions resulting from the electrochemical measurements made at the
ENEL facility include the following(44):

The EQUAT model is a reliable indicator of copper stability in the power plant
feedwater environment.

The lowest steady state feedwater copper concentrations (typically <1.5 ppb) were
attained with AVT.

Oxygen ingress while on AVT (due to condenser air inleakage, polisher vessel
changeouts, and, most significantly, startups) increases feedwater copper
concentrations and copper electrochemical potential.

OT, with either peroxide or oxygen, produced a partially protective layer of cupric
oxide on heater surfaces. Passivation was quicker with peroxide than oxygen.

Using OT, steady state copper levels ranged from 2-5 ppb and were generally at the
lower end of the range when conditioning with oxygen.

Loss of ammonia or oxidant levels while on OT results in lower pH and copper


electrochemical potential readings and can result in increased release of copper to
the feedwater.

The ENEL test program suggests that copper concentrations in feedwater can be
maintained at 4 ppb if chemistry is held in the following ranges while on OT.
pH @ 25C: 8.88.9
Specific Conductivity at 25C: 1.82.2 S/cm
4-33

EPRI Licensed Material


Electrochemistry of Copper

Oxidant: 60100 ppb as O2


Ammonia: 150200 ppb as NH3
However, steady state release rates for copper were consistently higher on OT than on
AVT and dont meet the Guiding Principles as outlined in Section 2. Also, the stability
of the copper oxide film formed during OT is unable to tolerate changes in chemistry;
iron oxide films on steel surfaces demonstrate much better resistance to short-term
excursions in oxidant level and pH.
Perhaps the most important result of this work is demonstration of monitoring
techniques using electrochemical principles to determine copper release and passivity
in a working fossil plant.
CONCLUDING REMARKS ON THE ELECTROCHEMISTRY OF COPPER ALLOYS IN
FEEDWATER
A significant deficiency in the understanding of the electrochemistry of copper arises
from the lack of elevated temperature potential-pH diagrams for the alloys of interest
(Cu-Zn, Cu-Ni) under feedwater conditions.
Also, there has only been rudimentary corrosion monitoring and measurement of the
oxidizing-reducing potential (ORP) of the feedwater despite the development by EPRI
and others of a whole suite of high temperature in-situ instruments, which could
provide direct information on feedwater material corrosion(41).
REFERENCES
1. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions , Pergamon Press,
New York, 1966.
2. M. Pourbaix. Atlas of Electrochemical Equilibrium in Aqueous Solutions . National
Association of Corrosion Engineers, (Now NACE International). Houston, Texas.
1974.
3. H. K. Arthur, L. B. Kriksunov, and D. D. Macdonald. The Electrochemistry of
Copper in Power Plant Environments International Conference on Interaction of NonIron Based Materials with Water and Steam . Piacenza, Italy. June 11-13, 1996.
4. M. O. Speidel. ARPA Handbook on Stress Corrosion Cracking .
5. Corrosion Related Failures in Feedwater Heaters . Electric Power Research Institute, Palo
Alto, Calif. EPRI CS-3184. June 1983.

4-34

EPRI Licensed Material


Electrochemistry of Copper

6. B. C. Syrett. A Review of Corrosion of Copper Alloys Exposed to High Purity


Waters. Interaction of Non-Iron Based Materials with Water and Steam . Edited by B.
Dooley and A. Bursik. EPRI Proceedings. TR-108236. July 1997.
7. Computer-Calculated Potential pH Diagrams to 300C, Volume 1: Executive Summary .
Electric Power Research Institute, Palo Alto, Calif. EPRI NP-3137, Vol. 1. June, 1983.
8. M. T. Froning and E. D. Verink, Jr. Annual Report for Office of Naval Research . 1976.
9. P. A. Brooks. Corrosion Science. Vol. II, p. 389, 1971.
10. P. Duby. Proceedings, High Temperature High Pressure Electrochemistry in Aqueous
Solutions. Edited by R. W. Staehle, D. de G. Jones, and J. E. Salter. NACE, 1976.
11. P. B. Linkson, B. D. Phillips and E. D. Rowles. Corrosion Science. Vol. 19, 1979.
12. Computer - Calculated Potential pH Diagrams to 300C, Vol. 2: Handbook of Diagrams .
Electric Power Research Institute, Palo Alto, Calif. EPRI NP-3137, Vol. 2, June, 1983.
13. Computer-Calculated Potential pH Diagrams to 300C, Volume 3: Users Guide for
Computer Program POT-pH-Temp . Electric Power Research Institute, Palo Alto, Calif.
EPRI NP-3137, Vol. 3, Chapter 9, June 1983.
14. The ASME Handbook on Water Technology for Thermal Power Systems . American
Society for Mechanical Engineers, New York, New York. Electric Power Research
Institute, Palo Alto, Calif. EPRI GS-6303. 1989.
15. D. D. Macdonald, G. Shierman and P. Butler, The Thermodynamics of Metal-Water
Systems at Elevated Temperatures. I. The Water and Copper-Water Systems.
AECL-4136 (1972). Atomic Energy of Canada Ltd.
16. C. Wagner and W. Traud. The Interpretation of Corrosion by Superimposition of
Electrochemical Partial Reactions and the Formation of Potentials of Mixed
Electrodes. Z. Elektrochem. 1938.
17. M. G. Fontana and N. D. Greene. Corrosion Engineering, First Edition . McGraw-Hill
Book Company, New York, New York. 1967.
18. C. Y. Chao, L. F. Lin and D. D. Macdonald. A Point Defect Model for Anodic
Passive Films. Journal of the Electrochemical Society . 1981.
19. L. F. Lin, C. Y. Chao and D. D. Macdonald. A Point Defect Model for Anodic
Passive Films II. Chemical Breakdown and Pit Initiation. Journal of the
Electrochemical Society. 1981.
20. A. J. de Bethune and N. A. S. Loud. Standard Aqueous Potentials and Coefficients at
25C. Clifford A. Hampel, Skokie, Illinois. 1964.
21. J. OM. Bockris. Parameters of Electrode Kinetics - Electrochemical Constants . NBS
Circular 524. U. S. Government Printing Office, Washington, D.C. 1953.
22. U. Bertocci and D. R. Turner, in Encyclo. Electrochem. Elements (Edit. A. J. Bard),
Marcel Dekker, New York, Vol. II, 1974.
4-35

EPRI Licensed Material


Electrochemistry of Copper

23. W. H. Smyrl, in Compr. Treat. Electrochem. (Edit J. OM. Bockris, B. E. Conway, E.


Yeager, and R. E. White), Plenum Press, New York, Vol. 4, 1981.
24. M. Vazquez, S. R. deSanchez, E. J. Calvo, and D. J. Schiffrin, J. Electroanal. Chem. 374,
179 (1994); 374, 189 (1994).
25. D. D. Macdonald. J. Electrochem. Soc. Vol. 121, 1974.
26. A. V. Manolescu, F. Chiarotto and P. Mayer. Failure of Low Pressure Feedwater
Tubing-A Case History. Presented at Corrosion 86. NACE, Houston, TX, 1986.
27. G. Brunoro, G. Perboni, G. Rocchini, and G. Trabanelli, Behaviour of Copper in
High Purity Water. Proc. of 11th International Corrosion Congress, Vol. 5.
Associazione Italiana Di Metallurgia, Milan, Italy. 1990.
28. M. Moliere, Y. Verdier, and C. Leymonie, Oxidation of Copper in High Purity
Water at 70C: Application to Electric Generator Operation. Corrosion Science. Vol.
30, No. 2/3. 1990.
29. D. Noel, A. Caramel, A. Le Gal La Salle, A. Jardy, R. Rosset, and M. Keddam.
Corrosion, Passivation and Protection of Copper in Aqueous Solutions, I: Cyclic
Corrosion Mechanism, Memoires et Etudes Scientifiques, Revue de Metallurgie.
March 1992. Also published as EDF Document HT-45/COM 1468-A, Electricite de
France, France. Sept. 1992.
30. Primer on Maintaining the Integrity of Water-Cooled Generator Stator Windings . Electric
Power Research Institute, Palo Alto, Calif. TR-105504. September 1995.
31. B. Bussert. GE Proprietary Internal Report. 1973. [Cited in Ref. 26]
32. D. D. Macdonald. J. Electrochem. Soc. Vol., 139, p. 3434, 1992.
33. D. D. Macdonald, B. C. Syrett, and S. S. Wing, Corrosion, Vol. 34, No. 9, September
1978.
34. C. B. Bargeron and R. B. Givens, Corrosion, Vol. 36, p. 618, 1980.
35. I. Milosev, M. Metikos-Hukovic, M. Drogowska, H. Menard, and L. Brossard, J.
Electrochem. Soc., Vol. 139, p. 2409, 1992.
36. H-H. Strehblow and B. Titze, Electrochim. Acta, Vol. 25, p. 339, 1980.
37. D. D. Macdonald and M. Urquidi, J. Electrochem. Soc., Vol. 132, p. 555, 1985.
38. B. C. Syrett. Reference Manual for On-Line Monitoring of Water Chemistry and
Corrosion. Electric Power Research Institute, Palo Alto, Calif. TR-104928. March
1995.
39. W. Y. Mok, W. M. Cox, P. S. N. Stokes, J. R. Chamberlain, and W. T. Bakker. OnLine Corrosion Surveillance in the Superheater Section of a Coal-Fried Power
Boiler. Presented at International Symposium on Improved Technology for Fossil
Power Plants - New and Retrofit Applications. Washington, DC, March 1993.
4-36

EPRI Licensed Material


Electrochemistry of Copper

40. High-Temperature Fireside Corrosion Monitoring in the Superheater Section of a


Pulverized-Coal-Fired Boiler. Electric Power Research Institute, Palo Alto, Calif.
TR-101799. March 1994.
41. Proceedings, 12th International Corrosion Conference, Vol. 6, Electric Power
Industry. NACE International, Houston, Texas. September 19-24, 1993.
42. W. Kassen, J. Seager and K. Beichel. Electrochemical Potential Monitoring in
Feedwater at the St. Lucie 2 PWR. (As published in Reference 41).
43. G. Magnani, G. Bianchi and R. Bisogni. Copper Behavior in Thermal Cycle Fluid
with Combined Water Treatment. Interaction of Non-Iron Based Materials with
Water and Steam. Edited by B. Dooley and A. Bursik. EPRI Proceedings. TR108236. July 1997.
44. G. Perboni, V. Scolori and R. Zoppetti. Electrochemical Monitoring of Copper
Release From Low Pressure Pre-Heaters. Interaction of Non-Iron Based Materials
with Water and Steam. Edited by B. Dooley and A. Bursik. EPRI Proceedings. TR108236. July 1997.
45. G. Perboni: Study on the Copper Solubility in the Physico-Chemical Conditions of
Low-Pressure Preheaters of Power Plants. Report ENEL/CRAM G6/94/02, April
1994 (in Italian).
46. R. B. Dooley and J. Mathews. The Current State of Knowledge of Cycle Chemistry
for Fossil Plants. Fifth International Conference on Fossil Plant Cycle Chemistry.
EPRI Proceedings TR-108459. December 1997.

4-37

EPRI Licensed Material

5
COPPER TRANSPORT IN FEEDWATER SYSTEMS

INTRODUCTION
Corrosion product transport studies have long been a preferred approach taken to
investigate and resolve problems involving feedwater corrosion and disposition of
oxides. The metals addressed by such evaluations naturally reflect the materials in use
in the unit in question. While assessments of all-ferrous cycles are limited to iron
transport, monitoring campaigns for mixed metallurgy units need to include copper
and other important alloying constituents such as nickel and zinc.
If properly conducted, corrosion product transport assessments can be beneficial to the
utility in many ways. Important facts which can be established include:

Sources and forms of corrosion products.

Locations subject to deposition of corrosion products.

Identification of chemical and/or operating conditions which influence release and


deposition of corrosion products.

Effects of changes in chemical treatment and/or operating practices on corrosion


product transport activity.

Significance of transport paths (carryover, attemperation water) by which corrosion


products enter the steam path.

Effectiveness of available corrosion product removal techniques (boiler blowdown,


wasting of heater drains, condensate demineralizers and filters, etc.)

Findings of transport studies also may be used to identify and prioritize future needs
with respect to visual inspections, nondestructive evaluations, specimen removal,
chemical cleaning and other maintenance activities.
Corrosion product monitoring is regarded as integral to efforts to successfully
commission new units and to modify and/or optimize chemistry in existing units(1-5).
Routine monitoring of corrosion products is rarely practiced at utility fossil plants due
to a combination of economic and technical factors. These most typically relate to the
inability to analyze samples for metals at the plant and the cost associated with having
such tests conducted elsewhere. However, even at those facilities where the necessary
5-1

EPRI Licensed Material


Copper Transport in Feedwater Systems

equipment and personnel expertise is available, metals testing has been de-emphasized
in many cases.
These trends appear to be driven by the industry-wide desire to improve the
competitive position by minimization of operating costs. They may be further
supported by the misconception that feedwater and boiler water treatments in use will
suffice to maintain corrosion product transport rates at acceptable levels. In the case of
fossil plant units where chemistry-related equipment damage, efficiency losses and
reduction in useful component life are experienced, it is likely that the treatment
program is deficient in one or more respects. Assessment of corrosion product
transport, as part of a comprehensive chemistry surveillance campaign, is often
beneficial in identifying areas of deficiency, which allows site-specific evaluation and
implementation of corrective actions.
The discussions of copper transport which follow are based on several published
accounts of experience at US and international fossil plants with mixed metallurgy
cycles which address observations with respect to behavior of copper alloys under
various service conditions(6-24). Overall, these field and laboratory studies tend to
substantiate the importance of the Guiding Principles (Section 2) in eliminating copper
fouling around the cycle.
Copper monitoring in the field represents a challenge to utility personnel requiring
reliable data. Problems encountered in plant studies, solutions applied by
investigators, and a summary of good practices are presented at the end of the section.
SOURCES OF COPPER
Available data from plant corrosion transport studies indicate that copper corrosion
and release may occur from any copper alloy component in contact with water or
steam. In addition, copper deposits formed under one set of chemistry and operating
conditions may reenter water or steam under other conditions. Also, copper deposited
onto all-ferrous HP heater tubes can, and has been, solubilized by a change to more
oxidizing conditions. In general, condenser tubes are not a significant source of copper.
Release from copper alloy high pressure heaters can be substantial, especially in units
with 70-30 copper-nickel tubes.
However, contributions from brass low pressure heaters are usually the most
significant as it is this region which has the poorest oxygen control due to air inleakage, particularly during peaking/cycling service during startup.

Copper Release from Low Pressure Heaters


Nearly all units for which results of copper transport studies have been published
featured brass tubes in low pressure heaters(610,1214,16,17, 2024). Low pressure units at one
5-2

EPRI Licensed Material


Copper Transport in Feedwater Systems

station were tubed with 90-10 copper-nickel(18). Units at another station employed 7030 copper-nickel in the low pressure heaters(19).
Release from brass heater tubes in low pressure heaters was variable and appeared to
depend on several factors. Some accounts point to heater steamsides as the primary
release point while others suggest both steamside and waterside release to be
significant. Differences in individual findings appear to reflect site-specific design
characteristics, operating requirements and chemistry treatments.
There are a number of reasons why the corrosion of copper and the release of copper
oxides can be high in the low pressure feedwater train:

High levels of oxygen (>20 ppb) and carbon dioxide due to poor air inleakage
control

Ineffectiveness of oxygen scavengers at the low temperatures in the LP feedwater.


It is necessary to keep the oxidizing-reducing potential (ORP) less than 0mV.

Reduction, elimination or change of the oxygen scavenger has led to numerous


instances of increased copper transport

None or ineffective shutdown and layup procedures to keep the ORP < 0mV.

Investigations which provide copper data at several points around the cycle support the
position that copper release by condenser tubes is normally minor in magnitude.
Appraisal of copper levels in condensate and low pressure heater drain samples clearly
show that most copper detected at the former location originates in the low pressure
heater shellsides(10).
Published accounts which assessed and reported on transport of zinc suggest release
rates of copper and zinc from brass heaters to be essentially stoichiometric(10). The
inference is that release of these metals occurs as a result of general corrosion activity.
This would appear consistent with research results indicating that rates of copper
corrosion and release in brass increase with temperature and accounts for the minimal
release of copper from condenser tubes as compared to low pressure heater shellsides.
Copper transport in units at the station where low pressure heaters were tubed with 9010 copper-nickel was attributed primarily to corrosion and release activity in the high
pressure heater tubes(18). As discussed in Section 3, 90-10 copper-nickel is more
corrosion-resistant than brass under equivalent service conditions. However, a 1996
survey of utilities focusing on turbine copper fouling indicated that units operated by
several respondees utilized both metallurgies in low pressure heaters, as well as 70-30
copper-nickel in a few instances(25).

5-3

EPRI Licensed Material


Copper Transport in Feedwater Systems

Copper Release from High Pressure Heaters


Copper alloys used in high pressure heaters of units for which metals transport
characteristics have been assessed included 70-30 copper-nickel(1214, 1618) and Monel(610,
15,19). Other references to utility corrosion product transport either involved steel tubed
high pressure heaters(2024) or did not specify the alloy(11).
Evaluation of copper transport data for units with copper alloy low pressure heaters
and 70-30 copper-nickel or Monel high pressure heaters reveals some important
transport characteristics which appear fairly consistent in all cases:

A significant fraction of the copper present in the feedwater deposits in the high
pressure heater system; this occurs mainly during periods of high-load service
operation. The deposition occurs on copper base alloys, Monel, carbon steel and
stainless steel tubing.

Copper release from high pressure heaters is experienced during transient operating
conditions; most significant among these is cold startups, however, release has been
reported at shutdown and after hot/warm restarts. In many units, feedwater
oxygen levels can be high because no steam is available for the deaerator during
early startup.

Release rates for copper and nickel from high pressure heaters is typically
nonstoichiometric and relative proportions of these metals in feedwater vary with
unit service conditions.

Copper release from alloys used in high pressure heaters appears to involve several
processes. Copper release from the waterside surfaces of tubing in the HP feedwater
heaters involves two processes:

general corrosion and release from the copper containing tubes

soluble and particulate release from the copper deposits

In the second case, where copper deposition is significant, periodic release of copperrich foils may occur (snakeskins). The general release of copper from deposits on allferrous materials has been observed on conversion to OT or elimination of the oxygen
scavenger. These changes result in oxidizing environments (ORP > 0 mV) in the HP
feedwater, which convert the copper deposits to cupric oxide.
Steamside releases are attributed primarily to exfoliation, a process which requires the
presence of oxygen. This process is exacerbated during layup and startup.
Nonstoichiometric release of copper and nickel from high pressure heaters may be a
consequence of dealloying activity. Metallurgical examination of 70-30 copper-nickel
tubes from units experiencing significant copper transport activity revealed indications
of denicklification(18). Transport data suggesting disproportionate release of nickel
5-4

EPRI Licensed Material


Copper Transport in Feedwater Systems

have lead other investigators to the same conclusion(17). Given the fact that copper is
more noble than nickel, the existence of some dealloying activity is not surprising.
Results of the various transport assessments generally support laboratory findings
which indicate Monel to be more corrosion-resistant than 70-30 copper-nickel under
high pressure heater service conditions. Most investigations of units with brass low
pressure heaters and Monel high pressure heaters have indicated that the low pressure
heaters are the primary source of copper. It would be incorrect, however, to conclude
that no release of copper occurs from Monel heaters. An assessment performed in the
late 1960s found indications of copper release from steamside surfaces of Monel
heaters(7). A study performed in the 1980s suggested that steamside releases of copper
from Monel heaters involved denicklification activity(10). Copper transport in a oncethrough boiler unit with Monel high pressure heaters (and no other identifiable source
of copper) was reported to be significant enough to result in turbine copper fouling(15).
Also, respondees to the 1996 utility survey on turbine copper fouling indicated Monel
and 70-30 copper-nickel to be the most widely used alloys in high pressure heaters.
Other copper-nickels (including 90-10 and 80-20 metallurgies) and stainless steel
(usually as a replacement material) were also reported by survey respondees(25).
FORM OF COPPER
In many instances, no effort was made to distinguish the presence of copper in solution
from that in suspension. Only total copper values were determined. Transport studies
reporting levels of copper present in dissolved and particulate forms typically passed
samples through a 0.45 micron filter and designated the portion of copper retained to
be present as particulate (suspended or insoluble) copper.
Findings reported with respect to the form in which copper was identified in water
samples from the preboiler indicate that both soluble an insoluble fractions are usually
present. However, there appears to be considerable variation in the relative amounts of
each observed. Brown and McSweeney of Ontario Hydro reported that copper was, at
times, equally represented in the soluble and colloidal states but was also determined
to be present in primarily the soluble form at other times(8,9). A study conducted by
Hook, Pearl and Sawochka for EPRI indicated that relative amounts of filterable and
nonfilterable copper were variable and could not be easily correlated with operating
conditions, chemistry, or sample point location(10). One example of the data at the LP
heaters and two HP heaters (#6 and 7) for copper as a function of the unit load is
shown in Figure 5-1. This figure illustrates most of the copper release coming from the
LP heaters and being deposited on the HP heaters. Both the release and deposition are
elevated immediately after a unit shutdown, with slightly higher levels of copper
release occurring from the LP heaters after a shutdown where the vacuum was broken.
Data for copper in boiler water included periods where both filterable and nonfilterable
copper represented the predominant form. Assessments of copper transport under
reducing and oxidizing conditions by Magnani, Bianchi and Bisogni(23) and by Perboni,
5-5

EPRI Licensed Material


Copper Transport in Feedwater Systems

Scolari and Zoppetti(24) suggest that soluble copper levels are higher under oxidizing
treatment than with reducing chemistry. Figure 5-2 shows that most of the copper is
present in dissolved form with oxygenated treatment(23). Iron, conversely, is present
almost entirely in suspended form.
Despite limited data on the form in which copper exists in fossil plant cycles, it would
appear that temperature, environment, materials and applicable release mechanisms all
influence the distribution. Effects of sample extraction and conditioning capabilities
probably represent a source of bias and effectively limit the comparability of findings
of individual investigations. Approaches taken in sample handling, particularly
separation of soluble and insoluble copper are also variable and may introduce
additional bias. However, for the purposes of evaluating and optimizing feedwater
treatment, determination of total copper values is generally regarded as sufficient.

Positive results indicate copper release, negative results indicate deposition.


Source: Reference 10.

Figure 5-1 Copper Concentrations at Port Everglades Unit 4 During July 1985

5-6

EPRI Licensed Material


Copper Transport in Feedwater Systems

Piombino Power Station, Unit 3


Filtration Results
5
4.5

Suspended Cu

Dissolved Cu

Copper, g/l

3.5
3
2.5
2
1.5
1

SH2O

SH1O

SH1I

ECOI

DEGO

DEGI

HPHD

LPHO

MBO

MBI

CPD

0.5

Sampling Points

Piombino Power Station, Unit 3


Filtration Results
5
4.5
Suspended Fe
4

Dissolved Fe

Iron, g/l

3.5
3
2.5
2
1.5
1

SH2O

SH1O

SH1I

ECOI

DEGO

DEGI

HPHD

LPHO

MBO

MBI

CPD

0.5

Sampling Points

Source: Reference 23.

Figure 5-2 Levels of Suspended and Dissolved Copper and Iron Present Under
Oxygenated Treatment

5-7

EPRI Licensed Material


Copper Transport in Feedwater Systems

EFFECT OF PEAKING AND CYCLING SERVICE ON COPPER TRANSPORT


It has been shown and is generally accepted that metals transport activity is higher
during startup and immediately after startup than during periods of sustained
operation. This was a fundamental conclusion of nearly every one of the utility plant
assessments cited, regardless of whether the units being evaluated were experiencing
problems related to deposition of copper or not. For example, Smith, in presenting
findings of a copper transport study performed to evaluate copper deposition in high
pressure heaters, noted that iron and copper levels increased at startup and remained
high for several days thereafter(6). Release of copper from the high pressure heaters
was also observed. Hagewood, Klein and Voyles noted that significant copper
corrosion took place during unit shutdowns and startups. This apparently was a
significant contributing factor to waterwall corrosion experienced after the boiler was
converted from baseload to peaking service(7). Copper transport was shown to be less
significant in similar boilers maintained in a baseloaded condition.
Figure 5-3 presents a plot of iron and copper concentrations in feedwater during
startup of a drum boiler unit in the Ontario Hydro System in the mid 1970s (8,9). The
initial peak is attributed to sloughing of oxides due to thermal expansion. The peaks
occurring approximately three hours after turbine rolling appear to correspond to
placement of additional condensate and boiler feed pumps in service in order to
increase steam flow. Despite these trends in metals transport, the authors judged
control to be satisfactory overall, and acknowledged the benefits of condensate
polishers in cleaning up the feedwater system. However, it is significant that the
elevated levels of copper (and iron) can exist for up to 56 hours after startup.
Surveys conducted during unit shutdown under the Ontario Hydro program indicated
only minor increases in metals concentrations relative to those evident during startup.
More recent investigations at other utilities present an opposite view, perhaps as a
result of differences in shutdown practices.
Hook, Pearl and Sawochka demonstrated that metals concentrations during unit
startup were largely influenced by the extent to which air entered the system during
the shutdown (10). Metals release was reduced through either maintaining condenser
vacuum during shutdown or isolating the heaters from the condenser (Figure 5-1).
Shells of low pressure heaters were shown to become air inleakage sources at
shutdown, as shown in Figure 5-4. Units subjected to load cycling may experience
similar effects if load is dropped low enough. The authors concluded that increased
oxygen levels resulting from air inleakage were more detrimental to iron transport than
copper transport. The effect of pH depression (via carbon dioxide) also may have
influenced the situation.

5-8

EPRI Licensed Material


Copper Transport in Feedwater Systems

Source: References 8 and 9.

Figure 5-3 Feedwater Metals Concentrations in Feedwater During Startup Operations

Figure 5-4 Relationship Between Power and Heater Shell Pressure


5-9

EPRI Licensed Material


Copper Transport in Feedwater Systems

Results of transport evaluations conducted in the last ten years bring the influence of
unit shutdowns and startups more clearly into focus. Hofmann reported detection of
copper levels in excess of one ppm in high pressure heater drains during shutdowns
and startups(13). Nitrogen blanketing of high pressure heater shells (to exclude oxygen)
and starting the unit up with the heaters in service instead of valving them in after
reaching minimum load (to minimize thermal shock and release of corrosion products)
were implemented to reduce copper release activity. As shown in Table 5-1, cycle
copper levels were reduced appreciably (although less than desired) as a result of
changes in these operating practices.
Evaluations of copper transport at a three-unit utility station by Sawochka, Clouse,
Larson and Archbold also indicate that release rates can be extremely high during unit
startup(17). Integrated corrosion product sampler devices were used during this work; a
schematic is shown in Figure 5-5.
Copper concentrations during normal unit operations were variable, and sometimes
exceeded the EPRI value of 2 ppb at the economizer inlet (Figure 5-6).

Table 5-1
Effect of High Pressure Heater Drains on System Copper Levels
Sample Location

May 1988

May 1992

Hotwell

627

330

7-1 Heater

504

7-1A LP Heater Drain

Sample Location

May 1988

May 1992

7-6B HP Heater Drain

9468

1480

260

7-7A HP Heater Drain

1326

1660

507

340

7-7B HP Heater Drain

11146

1930

Heating System
Returns

336

28

Deaerator

581

130

7-2 Heater Inlet

782

190

Economizer Inlet

1469

200

7-2 Heater Discharge

501

200

Boiler Drum

75300

5800

7-3 Heater Discharge

632

173

Saturated Steam

62.1

17.9

7-4 Heater Discharge

978

147

Superheated Steam

75.2

35.3

7-6A HP Heater Drain

4717

440

Note: All values in parts per billion. Both sets taken two hours after load on turbine, cold startup. May
1992 data reflect blanketing of heater shells and starting up with heaters in service.
Source: Reference 13.

5-10

EPRI Licensed Material


Copper Transport in Feedwater Systems

Figure 5-5 Corrosion Product Sampler Devices used at Laramie River Station

5-11

EPRI Licensed Material


Copper Transport in Feedwater Systems

Source: Reference 17.

Figure 5-6 Copper Concentrations During Normal Operation of Unit No. 2 at Laramie
River Station
Transport data indicated that copper was being released on the steamside of the high
pressure heaters and depositing on the waterside of the heaters. Nickel release rates
were nonstoichiometric, an indication of denicklification activity.
Data for transport during normal service operation were accepted as a reasonable
explanation of copper deposition in the feedwater system. This had reportedly
manifested itself as fouling of boiler feed pump strainers and pluggage of feedwater
5-12

EPRI Licensed Material


Copper Transport in Feedwater Systems

heater tubes(16). Copper concentrations reported at the economizer inlet did not,
however, appear substantial enough for the levels of copper determined present on
boiler waterwalls. Chemical cleaning of the boilers were conducted after deposit
densities in the range of 13 to 28 grams per square foot were reported. (Copper oxide
levels on the order of 61 to 80 percent and only 12 to 14 percent iron oxide were
present.)
Chemical cleaning of the three boilers resulted in removal of 7508 pounds of copper
from Unit 1, 2616 pounds from Unit 2 and 2966 pounds from Unit 3. Given this
inventory of copper it is not surprising that deposition in the superheaters and turbines
also had been experienced(16,17).
Monitoring corrosion product transport during a shutdown and ensuing startup of one
of the units revealed extremely high copper values initially in feedwater at the
economizer inlet (Figure 5-7). Copper transport over the first five days of operation
was 10 to 20 times that detected during normal unit operations. It was concluded that
copper transport during this time period was equivalent to that resulting from two
months of normal operation. Copper release during the shutdown was not quantified
but was indicated to be much less significant than at startup.
The literature includes other, although less detailed accounts of substantial copper
transport. Lawrence reports that fouling of turbines at the Columbia Energy Center is
largely attributable to denicklification of 70-30 copper-nickel high pressure heaters (18).
Most of the copper release was determined to occur during major transients. Copper
levels at startup were noted to be 60 to 70 times greater than during normal operations.
Devalois, Gilchrist and Price noted that substantially higher corrosion products were
present between the time the unit was placed in service and was released to full load
operation(19). As a consequence, chemical cleaning of the boilers and turbines of the two
units with 70-30 copper-nickel low pressure heaters and Monel high pressure heaters
was required. A sister unit, equipped with stainless steel feedwater heaters and a
condensate polisher was subject to less significant excursions in corrosion products.
Using these findings, the station was able to develop an economic justification to
retrofit a condensate polisher to be used for startup of the units with copper feedwater
heaters.

5-13

EPRI Licensed Material


Copper Transport in Feedwater Systems

Figure 5-7 Corrosion Product Transport During Return to Service of Unit No. 1 at
Laramie River Station (April 1522, 1991)
MINIMIZING COPPER TRANSPORT
Thermodynamically it is not possible to eliminate all copper transport in mixed
metallurgy feedwater systems. The two primary oxides, Cu2O and CuO, and
hydroxide Cu(OH)2 grow as corrosion products on the surface of copper-based
feedwater tubes as discussed in Section 3. These eventually become incorporated into
the feedwater. As indicated in the Guiding Principles in Section 2, it is, however,
possible to control the copper at the economizer inlet sampling point to less than the
5-14

EPRI Licensed Material


Copper Transport in Feedwater Systems

EPRI Guideline value of 2 ppb. The Guiding Principles also provide guidance on those
operating and chemistry features which will minimize copper problems.
Approaches traditionally suggested to minimize problems associated with copper
transport in feedwater systems typically employ two strategies. The first, which is
much preferred, involves measures intended to reduce corrosion of copper alloys and
release of copper oxides to the feedwater system. The second involves removal of
released copper oxides from the feedwater before it is transported to locations at which
deposition is likely to occur.
Table 5-2 summarizes complementary actions which support the Guiding Principles.
With regards to keeping a reducing environment in the feedwater during all periods of
operation and shutdown, the first requirement is to employ and optimize a reducing
agent (oxygen scavenger, such as hydrazine). This aspect is covered in Section 8. The
second is to minimize oxygen through air in-leakage control during operations, by
blanketing heaters during shutdown, and by starting up the unit with deoxygenated
water.

Table 5-2
Mitigation of Copper Transport
Technique

Purpose

Application of scavenger

Production of reducing environment

Blanketing of heaters and boilers

Oxygen exclusion during shutdown

Air inleakage control

Oxygen exclusion

Startup with deoxygenated water

Oxygen exclusion

Boiler drum level control during


startup/transients

Minimize copper carryover

Optimize heater procedures

Minimize thermal shock and copper


spalling; oxygen exclusion

Condensate polishers

Copper removal

Retube heaters

Copper elimination

Cleanup loop

Reduce copper

Condensate polishers have been proven to be very effective devices for removing
copper from the preboiler system. In most plant designs, polishers are only able to
treat the condensate, which includes any copper released by the condenser and low
pressure heater shellsides. Many units with mixed metallurgies do not include
polishers. Space limitations and capital costs represent barriers to polisher retrofits,
5-15

EPRI Licensed Material


Copper Transport in Feedwater Systems

however, in cases where copper transport is causing turbine fouling, it may be possible
to determine a favorable benefit/cost analysis for a system sized to handle unit
startups(3133).
Boiler blowdown is also capable of removing some copper from the steam-water cycle.
However, based on findings of the published transport studies, it is concluded that,
generally, this approach provides only limited benefit. As discussed in Section 6,
blowdown control may be of some use during startup to minimize transport of copper
from the boiler water to the steam path. Wasting of heater drains at startup also was
identified as a useful practice by some investigators but design and operating concerns
typically restrict the extent to which this practice - which also represents an energy
penalty to the unit - can be applied.
Consistent with the conclusions of laboratory investigations of corrosion of materials in
the feedwater environment, the copper alloy heater tubes should ultimately be replaced
with copper-free materials such as 300 series stainless steels or low chromium
containing ferritics such as 2 1/4Cr1Mo (Section 8). This is an especially important
consideration at those plants where release from high pressure heaters is significant
and removal by condensate polishers (where provided) is not feasible unless a
condensate/feedwater cleanup loop can be established(26). A schematic arrangement is
shown in Figure 5-8(5).
Optimization of heater operating procedures is often integral to oxygen exclusion but,
more importantly, can reduce the thermal shock which causes spalling of copper-rich
material from the surfaces.
MONITORING COPPER TRANSPORT ACTIVITY
In view of the findings of various copper transport assessments previously presented,
the desirability of monitoring copper levels in all mixed metallurgy units seems
apparent. (Guidance is provided in Section 8.) Based on the experience of laboratory
and field investigations, utility personnel may need to evaluate and upgrade sample
extraction and conditioning systems as well as select a sample collection method and
analytical technique. Only those points which need to be made with respect to copper
follow, since more comprehensive information on this subject is available elsewhere(27
29).
An excellent evaluation of issues related to copper sampling in the power plant is
provided by Sigon and Balconi(22). Collection of a representative sample is the initial
concern. Sample nozzles which maintain integrity with respect to relative amounts of
dissolved and particulate copper species should be used. Considerable effort has been
expanded in evaluation of appropriate nozzles for isokinetic sampling of high pressure
feedwater and steam(30).
5-16

EPRI Licensed Material


Copper Transport in Feedwater Systems

(Dashed lines denote cleanup loops for low and high pressure heaters)
Source: Reference 5

Figure 5-8 Copper, Nickel and Zinc Generation, Transport and Effects in a Drum Boiler
Unit with Mixed Metallurgy
Assuming that a representative sample with respect to copper can be attained, attention
must then be directed to the sample conditioning system. Changes in sample pH and
copper solubility resulting from temperature reduction create conditions which favor
copper deposition within sample lines. Under typical AVT conditions (<200 ppb
ammonia), the solubility of copper in water at 25C is estimated to be < 2 ppb(22). This
would seem to suggest that some of the soluble copper values reported in the literature

5-17

EPRI Licensed Material


Copper Transport in Feedwater Systems

included fine particulates which could pass through the filters used to separate
filterable and nonfilterable fractions.
Minimization of copper deposition is practically accomplished by locating primary
sample coolers as close to the extraction point as possible and collecting the sample at
as close to the cooler outlet as practical. Use of in-line filters should be avoided as they
represent a site for capture of particulate copper. Also, in-line filters have been
observed to serve as a reaction site for oxidants, reductants and metals. Given the
effects of dissolved oxygen and hydrazine on electrochemical potential, use of filters
would influence any reduction-oxidation potential measurements as well as the
concentrations of oxygen and/or hydrazine present downstream of the filters.
Copper analysis may be made by means of laboratory analysis of grab samples, by
laboratory analysis of filter pads and ion exchange media from corrosion productsamplers, or by on-line monitoring methods. Laboratory methods available for copper
analysis include atomic absorption inductively coupled plasma and ion
chromatography. Anodic stripping voltammetry has also been used to monitor copper
at low levels in power plant samples(21). Analysis of particulate copper species may be
made by x-ray diffraction if it is desired to identify the actual copper compounds
present. On-line monitors employing colorimetric techniques(22, 29) and ion
chromatography(22) are also available.
Clearly, the integrated corrosion product samplers appear to be a better means of
collecting samples for metals analysis than grab sample collection. This is due to the
fact that the collection technique includes all dissolved and suspended metals collected
over the time interval the device is in use. Aliquots taken from individual grab
samples/or composite samples are subject to variable levels of particulates and,
therefore total metals. Sample conditioning requirements for integrated samplers are
less extensive than those for on-line analyzers, which may be useful when attempting
to assess the influence of plant operations and changes in chemistry on copper
transport trends in real time.
REFERENCES
1. Cycle Chemistry Guidelines for Fossil Plants: All Volatile Treatment. Electric Power
Research Institute, Palo Alto, Calif. EPRI TR-105041. April 1996.
2. Cycle Chemistry Guidelines for Fossil Plants: Oxygenated Treatment. Electric Power
Research Institute, Palo Alto, Calif. EPRI TR-102285. December 1994.
3. Cycle Chemistry Guidelines for Fossil Plants: Phosphate Treatment for Drum Units.
Electric Power Research Institute, Palo Alto, Calif. EPRI TR-103665. Final Report,
December 1994.
4. Sodium Hydroxide for Conditioning the Boiler Water of Drum-Type Boilers. Electric
Power Research Institute, Palo Alto, Calif. TR-104007. January 1995.
5-18

EPRI Licensed Material


Copper Transport in Feedwater Systems

5. Selection and Optimization of Boiler Water and Feedwater Treatments for Fossil Plants.
Electric Power Research Institute, Palo Alto, Calif. TR-105040.
6. O. A. Smith A Program to Locate Copper Pickup in a High-Pressure Utility
System. Proceedings of the American Power Conference, Volume 29. 1967.
7. B. T. Hagewood, H. A. Klein and D. E. Voyles. The Control of Internal Corrosion
in High-Pressure Peaking Units. Proceedings of the American Power Conference,
Volume 30. 1968.
8. J. Brown and P. McSweeney. Feedwater Line Corrosion. Proceedings of the
American Power Conference, Volume 39. 1977.
9. J. Brown and P. McSweeney. Feedwater Line Corrosion, Combustion, Volume 49,
No. 2, August 1977.
10. Corrosion-Product Transport in a Cycling Fossil Plant. Electric Power Research
Institute, Palo Alto, Calif. EPRI CS-5033. February 1987.
11. B. T. Hagewood CuO Fouling Northport 4 High-Pressure Steam Turbine. Second
Conference on Fossil Plant Cycle Chemistry. Electric Power Research Institute, Palo
Alto, Calif. August 1988.
12. M. L. Hofmann. Analysis of Copper Oxide in Two Steam Generators. Ninth
Annual Electric Utility Workshop. Champaign Illinois, 1989.
13. M. L. Hofmann. Copper Transport at Miami Fort Station. Fourteenth Annual
Electric Utility Workshop. Champaign, Illinois, 1994.
14. M. L. Hofmann. Concerted Utility Effort Tames Copper Deposition. Power.
Volume 138, No. 6, June 1994.
15. R. G. Axley, D. W. Beavers, S. O. Hilton and M. G. Sexton. Turbine Copper
Deposits in a Supercritical Once-Through Unit. Proceedings: International
Conference on Fossil Plant Cycle Chemistry. Electric Power Research Institute, Palo
Alto, Calif. EPRI TR-100195. December 1991.
16. T. Fitzsimmons, B. Larson, S. J. Arrington and L. Slabodnik. Copper Problems at
Laramie River Station. Proceedings: 52nd Annual Meeting, International Water
Conference. Engineers Society of Western Pennsylvania. 1991.
17. S. G. Sawochka, M. E. Clouse, B. Larson and T. Archbold. Corrosion Product
Control at the Laramie River Station. Proceedings: 53rd Annual Meeting, International
Water Conference. Engineers Society of Western Pennsylvania. 1992.
18. G. S. Lawrence. Chemical Cleaning of HP Turbines at Columbia Energy Center.
Proceedings: 56th Annual Meeting, International Water Conference. Engineers Society of
Western, Pennsylvania. 1995.
19. R. Devalois, T. Gilchrist and K. Price. Justification of the Retrofit of a Condensate
Polishing System at Tri-State G&Ts Craig Station. Presented at EPRI Condensate
Polishing Workshop: Deep Bed and Powdered Resin Systems. New Orleans, LA,
September 1517, 1993.
5-19

EPRI Licensed Material


Copper Transport in Feedwater Systems

20. D. Gunn, D. McInnes, K. Knights and D. Ryan. Combined Oxygen Hydrazine


Ammonia Conditioning (COHAC) and Copper Transport Through the Unit Cycle.
Interaction of Non-Iron Based Materials with Water and Steam. Edited by B. Dooley and
A. Bursik. EPRI Proceedings. TR-108236. July 1997.
21. K. N. Thomsen and K. Daucik. Deposits on Turbines in Relation to the Use of
Copper Alloy Materials in the Water/Steam Cycle. Interaction of Non-Iron Based
Materials with Water and Steam. Edited by B. Dooley and A. Bursik. EPRI
Proceedings. TR-108236. July 1997.
22. F. Sigon and M. L. Balconi Copper Species Monitoring in the Water Steam Cycle of
Power Plants by Ion Chromatography. Interaction of Non-Iron Based Materials with
Water and Steam. Edited by B. Dooley and A. Bursik. EPRI Proceedings. TR-108236.
July 1997.
23. G. Magnani, G. Bianchi and R. Bisogni. Copper Behavior in Thermal Cycle Fluid
with Combined Water Treatments. Interaction of Non-Iron Based Materials with Water
and Steam. Edited by B. Dooley and A. Bursik. EPRI Proceedings. TR-108236. July
1997.
24. G. Perboni, V. Scolori and R. Zoppetti. Electrochemical Monitoring of Copper
Release From Low Pressure Pre-Heaters. Interaction of Non-Iron Based Materials with
Water and Steam. Edited by B. Dooley and A. Bursik. EPRI Proceedings. TR-108236.
July 1997.
25. Turbine Copper Fouling Survey Database. 1996 (Unpublished).
26. D. M. Sopocy and Y. H. Lee. Corrosion Product Control Measures For Existing
Plants. Proceedings of the American Power Conference, Volume 47. 1985.
27. Guidance Manual on Instrumentation and Control for Fossil Plant Cycle Chemistry.
Electric Power Research Institute, Palo Alto, Calif. EPRI CS-5164, April 1987.
28. Monitoring Cycle Water Chemistry in Fossil Plants, Volume 3: Project Conclusions and
Recommendations. Electric Power Research Institute, Palo Alto, Calif. EPRI CS-7556,
Vol. 3, October 1991.
29. Reference Manual for On-Line Monitoring of Water Chemistry and Corrosion. Electric
Power Research Institute, Palo Alto, Calif. TR-104928. March 1995.
30. Development of a Steam Sampling System. Electric Power Research Institute, Palo Alto,
Calif. EPRI TR-100196, December 1991.
31. Condensate Polishing Guidelines. Electric Power Research Institute, Palo Alto, Calif.
EPRI TR-104422. September 1996.
32. Cycle Chemistry Improvement Program . Electric Power Research Institute. Palo Alto,
Calif. EPRI TR-106371. April 1997.
33. M. A. Sadler and R. B. Dooley. Possible Methods of Reducing the Cost of
Condensate Polishing on Fossil Plants. Proceedings Fifth International Conference on
Fossil Plant Cycle Chemistry . EPRI TR-108459. December 1997.
5-20

EPRI Licensed Material

6
VOLATILITY AND SOLUBILITY OF COPPER IN STEAM

INTRODUCTION
Previous chapters have described the corrosion, electrochemistry and transport of
copper and its oxides in the feedwater system. The copper oxides generated in the
feedwater are transported into the boiler, and can be incorporated into the steam where
they may deposit and be reduced. Analysis of the deposits through the various turbine
stages presented in terms of weight percent versus specific volume (inverse pressure or
density) indicates cupric oxide, (tenorite, CuO) and cuprous oxide (cuprite, Cu2O) to be
predominant under the low specific volume conditions encountered in the HP turbine.
Transport studies generally indicate carryover in drum boiler units to be a significant
and usually the primary path by which copper enters the steam path. This section
presents available information pertaining to mechanical and vaporous carryover of
copper and its oxides. Also, the limited data addressing the solubility of copper and its
oxides in steam is discussed.
CARRYOVER IN DRUM BOILERS
In most drum boiler units, carryover of solids from the boiler water to the steam
usually accounts for the majority of contaminants present. Effects of attemperation
spray water on steam purity can be assessed by comparison of impurity levels in
saturated and main/reheat steam samples or by calculation if the feedwater
composition, spray rate, saturated steam composition and boiler steaming rate are
known(1).
Carryover is defined as the ratio of the concentration of a chemical constituent in
saturated steam to the concentration of the constituent in the boiler water(2). All forms
in which the constituent exists (molecular, ionic, elemental) are considered. Total
carryover is the sum of mechanical and volatile carryover for the constituent of interest.

Mechanical Carryover
Mechanical carryover occurs as the result of impurities present in the liquid phase (the
moisture content) of the saturated steam. The rate of mechanical carryover is
influenced by the design and integrity of boiler steam drum internals, specifically the
moisture separator devices and baffles. Mechanical carryover also may be influenced
6-1

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

by boiler operating conditions, particularly drum level control, and particularly during
startup and transient conditions. Mechanical carryover increases with boiler pressure;
the typical values which are used for deriving guideline limits for properly designed
boilers with the internals in good condition are indicated in Figure 6-1(3). It should be
noted that a safety factor of 2 is included in this figure to provide a conservative
estimate of mechanical carryover(1). Actual mechanical carryover rates should be
measured when trying to identify the root causes of copper in steam. Boiler water
dissolved solids levels and composition will also influence mechanical carryover and
steam purity. The full procedure to determine mechanical carryover is described
within the Cycle Chemistry Improvement Program (CCIP)(2).

Figure 6-1 Representative Drum Boiler Mechanical Carryover (3)

6-2

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

Volatile Carryover
Some chemical species in solution actually undergo partial vaporization in the boiler
and exist in saturated steam in the vapor phase. Volatile carryover is the term used to
describe the extent to which chemical constituents are present in the saturated vapor.
The volatility of silica served as a barrier to high pressure unit designs until ion
exchange technology capable of reducing silica in the treated makeup to very low levels
became available. As with silica, copper volatility is significant in high pressure drum
boiler units.
The ray diagram, shown in Figure 6-2, provides an estimate of vaporous carryover
for several chemical species which may be present in the boiler water of units with
drum boilers(3). Inspection of the diagram indicates that predicted volatile carryover of
cuprous oxide is very close to that of silica. Cupric oxide exhibits a greater tendency
for volatilization than either silica or cuprous oxide; at 26002800 psi the volative
carryover of cupric oxide is between 3040% of the concentration in the boiler drum.
The higher predicted volatility of copper in the cupric state is consistent with
theoretical assessments and field experiences covered in prior sections of this report
and fully supports one of the Guiding Principles in Section 2 that copper transport to
the steam path can be reduced through provision of reducing conditions in the
feedwater, which favor the formation of cuprous oxide.

Carryover Testing
Carryover testing is generally accepted as integral to efforts to establish appropriate
chemistry programs for fossil units with drum boilers(3-6). Equipment manufacturer
guarantees and testing protocols for carryover are normally based on analysis of
sodium in the boiler blowdown and saturated steam. Data included in the ray
diagram (Figure 6-2) clearly indicate that sodium salts (e.g. phosphate, chloride,
sulfate and hydroxide) exhibit significantly lower volatility than copper oxides and
silica. The difference is at least an order of magnitude at 2800 psig drum pressure and
increases at lower pressures.
Utility assessments of copper carryover activity need to consider both mechanical and
volatile transport. Mechanical carryover (based on sodium testing) is necessary to
evaluate boiler performance relative to design specifications and performance
guarantees. This is especially important in the case of units with boilers on solidsbased treatment programs which experience deposition of sodium compounds in the
high pressure turbine. It is also important for controlled circulation boilers and designs
which are subject to crack formation in drum liners and baffles. Volatile carryover of
copper also needs to be appraised in order to assess the extent of volatile transport and
to determine the effect of available corrective actions on introduction of copper to the
steam path.
6-3

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

Source: Reference 7, 8

Figure 6-2 Distribution Ratios for Common Boiler Water Contaminants


VOLATILITY OF COPPER IN DRUM BOILERS
Work performed in Russia to develop the ray diagram data involved laboratory
testing and measurements(7, 8). The water used in the evaluations was generally
prepared with higher concentrations of the species of interest than would exist in
operating power plants. Solutions used were generally free of other chemical
constituents except the one to be evaluated during the experiment.
Limitations of the ray diagram were subsequently confirmed during a program of
chemistry monitoring at selected power stations in the United States(9, 10). These
assessments showed that concentrations of chloride and sulfate in steam could be as
much as two orders of magnitude greater than predicted by the ray diagram. To
improve the understanding of the volatility of salts, acids and bases in steam. EPRI
6-4

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

research has led to an understanding of the partitioning constants discussed in Section


1 and shown in Figure 1-2(1113).
In support of Program Copper, EPRI is continuing the study to investigate the volatility
of copper. The initial task in this effort involved an assessment of the applicability and
reliability of available data on the thermodynamics of copper/copper oxide solubility,
hydrolysis and complexation, solubility in steam, influence of complexing species on
copper solubility in steam and surface adsorption effects. Some of the important
findings and conclusions were summarized by Palmer(14):

Cupric oxide appears to be the most stable solid phase present in fossil plant cycles.

Cupric oxide is more soluble in steam than either cuprous oxide or metallic copper.

Solubility of copper oxides generally increases with temperature and steam density.

Copper carbonate (CuCO3) may be able to enhance the volatility of cupric oxide.

Cuprous chloride (CuCl) may enhance the solubility of cuprous oxide.

Cuprous chloride complexes could exhibit high volatility.

Copper-ammonia complexes are less volatile than copper oxides but may enhance
the solubility of cupric oxide.

Adsorption of cupric ion on magnetite is strong under ambient conditions and


should increase with temperature; coordination with ammonia may inhibit
adsorption.

Feedwater oxygen control is essential in order to minimize rates of oxidation of


copper and formation of volatile copper species.

This research will continue at ORNL to better define copper and copper oxide
volatility, and to assess the effects of other species which are present under power plant
conditions.
Previous investigations conducted in Russia have evaluated deposition of copper(15),
solubility of copper in supercritical units(16), copper deposition(17) and copper
distribution ratio(18). Review of this work by Allmon(19) provides the following points:

Soluble copper ammonia complex ions formed in the feedwater begin to break
drown at temperatures of 110150C (230302F). Copper oxides formed as a result
of this process are less soluble and can form deposits(15).

Assessment of copper in the superheater of supercritical units(16) indicated that


deposit formation occurs in the temperature range of 600-650C (11001200F).

These observations appear to be consistent with findings of copper transport studies


conducted at domestic fossil plants and reports of copper deposits in high pressure
heaters and boilers.
6-5

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

In drum boiler units, the distribution ratio for copper is reduced in the presence of
ammonia (17, 18); the concentration of hydrated copper oxide is reduced from that
which would be present in the absence of ammonia(19).

This last point was reaffirmed in a recent paper(20) where evaluations were conducted in
Russia to assess the solubility of copper in water and the distribution ratio at pressures
of 14 MPa (2030 psia) and 18 MPa (2610 psia). Results are given in Table 6-1. Solubility
experiments were conducted with copper concentrations of 5300 ppb. Solubility was
fairly stable over the pH range of 6.59.7 but increased with further pH depression;
distribution ratio was also stable over this range but decreased with increased copper
solubility. These changes are assumed to reflect formation of ionic copper species.
Analysis of insoluble copper collected on filter pads during the experiments
determined it to be present primarily as cupric oxide.

Table 6-1
Solubility of Copper in Water and Distribution Ratio
Test Pressure, MPa (psia)

Solubility, ppb

Distribution Ratio *

14 (2030)

4.50

0.10

18 (2610)

6.50

0.19

* Distribution ratio = copper concentration in steam/copper concentration in water


Source: Reference 20.

Subsequent experiments examined the effect of complexing agent (ammonia and


disodium EDTA) additions on distribution ratio. The ammonia experiments were
conducted at 18 MPa (2610 psia) with copper concentrations of 30 50 ppb and
ammonia concentrations of 0 3.5 ppm. Some of the analytical data from these
experiments are presented in Table 6-2. The presence of ammonia results in significant
reduction of the distribution ratios. This effect is most significant at ammonia levels of
1 2 ppm. Further increases in ammonia led to higher distribution ratios. The
investigators assume that further increases in ammonia level result in formation of
copper ammonia complexes of higher volatility.
Table 6-2
Influence of Ammonia on Distribution Ratios for Copper Oxide

6-6

Ammonia Concentration,
ppm NH4 OH

pH

Copper, ppb

Distribution
Ratio

0.14

8.30

Steam
1.60

Water
36.0

0.044

0.16

8.35

1.30

27.5

0.047

0.26

8.55

0.7

19.0

0.036

0.37

8.65

0.7

19.2

0.036

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

Table 6-2
Influence of Ammonia on Distribution Ratios for Copper Oxide
0.43

8.75

0.5

12.7

0.038

0.72

8.96

0.3

12.3

0.024

0.95

9.38

0.2

9.0

0.022

0.96

9.26

0.2

9.80

0.020

1.29

9.45

0.4

17.7

0.023

1.60

9.45

0.1

6.7

0.015

1.65

9.55

0.2

12.5

0.016

1.80

9.67

0.3

13.0

0.023

1.90

9.65

0.1

6.2

0.016

2.10

9.69

0.3

14.5

0.021

2.40

9.70

0.4

13.0

0.030

2.43

9.60

0.2

5.30

0.037

2.85

9.75

0.4

11.0

0.036

3.19

9.75

0.9

18.0

0.050

3.34

9.88

0.7

12.6

0.055

Source: Reference 20.

The ammonia levels at which reductions in distribution ratio are minimized are
impractical for fossil plant service, however, even the low levels of ammonia typically
present in the boiler water of mixed metallurgy cycles would appear to reduce copper
volatility relative to volatility in ammonia-free water. In view of the Russian
experimental data, it might also appear that the benefits associated with increased
feedwater hydrazine residuals could involve volatility effects (via ammonia formed
during thermal decomposition) as well as minimization of copper release and cupric
oxide formation.

Field Experience
At least one utility has implemented changes in chemistry and operations intended to
reduce the volatility of copper in the boiler water(21). The approach followed involves
maintaining reducing chemistry conditions by keeping the feedwater hydrazine
residuals high. The intent is to maximize the amount of copper oxide present in the
less volatile form (cuprite, Cu2O) and minimize concentrations of the more volatile
form (tenorite, CuO).
During startup of susceptible units at the Miami Fort Station, a boiler water copper
concentration/ drum boiler pressure control curve is utilized to minimize volatile
carryover activity. The control curve, based on actual operating data and plant
experience, has been revised several times; two versions are shown in Figure 6-3.
6-7

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

Copper control curves are used in the same fashion as a silica control curve is normally
applied at start-up with pressure restrictions (copper holds) enforced as needed to
avoid excessive volatile carryover. Blowdown is practiced as required to help
minimize startup times.

Figure 6-3 Boiler Water Copper Concentration/Drum Boiler Pressure Control Curves
Developed at Miami Fort Station (21)

SOLUBILITY OF COPPER IN STEAM


The predominant mechanism for copper carryover into steam appears to be related to
the solubility of cupric and cuprous oxides in steam.
The only experimental data for the solubility of copper and cuprous oxide in steam
were derived in the pioneering investigation by Pocock and Stewart(22). These
investigators also derived data for cupric oxide; three other investigations have also
been performed on this oxide(23,24,25).
6-8

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

The tests of Pocock and Stewart were conducted in autoclaves, using metallic copper,
cuprous oxide and cupric oxide. Blank runs were conducted to determine if copper
would be picked up from stainless steel materials in use (0.18 percent copper) or as a
consequence of chemical cleaning the sample coils (with nitric acid) between runs.
Most experiments were conducted at 1150F, 4500 psig and pH levels of 7.5 (untreated
demineralized water) and 9.5 (ammonia treated demineralized water).
The solubilities were found to be a strong function of pressure, Figure 6-4, or of the
specific volume of steam, but are relatively independent of temperature, at least over
the range 9001150F (482621C). In particular, the solubility for CuO increases
sharply above 15.8 MPa (2300 psi) and for Cu2O above 17.2 MPa (2500 psi), with the
CuO being more soluble. Cupric oxide (CuO) has about twice the solubility of cuprous
oxide (Cu2O) at 17.2 MPa (2500 psi), which is, in turn, more soluble than Cu in steam.
The solubility of CuO was greater at higher pH levels (9.5 vs. 7.5) (Table 6-3), a
behavior attributed to the greater ease with which CuO dissolves in solutions
containing ammonia.
Hearn et al(25) also measured the solubility of CuO from 57551C (1341022F) at
moderate pressures ranging from 76419 bar (11006070 psi) and found solubilities
almost on order of magnitude higher than those of Pocock and Stewart(22).

Table 6-3
Maximum Copper Solubilization at 4500 psi, 1150F
Solubility, ppb
Species

at pH 7.5

at pH 9.5

Metallic Copper

6.3

6.3

Cuprous Oxide

8.9

10.2

Cupric Oxide

15.4

17.2

Source: Reference 22

These discrepancies clearly need to be resolved. Some additional comments can be


made with respect to the experiments of Pocock and Stewart. The first involves
observations of copper deposition in the cooling coils used to condense steam
generated in the autoclave. As much as two thirds of the copper solubilized in the
autoclave was recovered during cleanings. Similar problems may be encountered
when attempting to collect high temperature samples for copper analyses in the field.
The second comment is related to efforts to avoid mechanical carryover of suspended
copper during the test runs; with the exception of three tests, all copper in the
condensed steam was reported to be present in dissolved form. However, carryover of
6-9

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

particulate copper (and magnetite particles to which copper has adsorbed) may occur
in an operating boiler.
However, there are at least four important conclusions from the data of Pocock and
Stewart which help in understanding the mechanism of copper transport and
deposition, and which fully support the Guiding Principles in Section 2:

First, since the highest state of oxidation (cupric oxide) has the highest steam
solubility, then maintaining reducing conditions around the cycle, especially in the
feedwater during operation and shutdown, and in the boiler water during
shutdown, will minimize the transport into the steam.

Second, the predominant deposited oxide in the turbine often is tenorite (cupric
oxide) which indicates that oxidizing conditions have occurred and assisted in the
transport of copper to the turbine.

Third, there is very little steam solubility for either copper oxide specie below the
range of 15.817.2 MPa (23002500 psi), which explains the field observations of
very little problem with copper deposition in lower pressure (<15.8 MPa (2300 psi))
units.

Fourth, the solubility of copper oxides increases with increasing pressure,


explaining why the units operating above 2500 psi (17.2 MPa) are most susceptible
to copper fouling.

CONCLUDING REMARKS
It is clear that a much better understanding of the solubilities of the main oxides, Cu2O
and CuO, is needed across the temperature and pressure range of interest. Also the
effects of ammonia and other complexing ligands (carbonates, chlorides and
phosphates) need to be investigated, as many compounds have been found in turbine
deposits (Section 7).

6-10

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

Figure 6-4 Solubility of cuprous (Cu2O) and cupric (CuO) oxides in superheated steam
as a function of pressure. Abstracted from data by Pocock and Stewart(22).
REFERENCES
1. Interim Consensus Guidelines on Fossil Plant Cycle Chemistry . Electric Power Research
Institute, Palo Alto, Calif. EPRI CS-4629. June 1986.
2. Cycle Chemistry Corrosion and Deposition: Correction, Prevention and Control. Electric
Power Research Institute, Palo Alto, Calif. TR-103038. Palo Alto, Calif., December
1993.
3. Cycle Chemistry Guidelines for Fossil Plants: All-Volatile Treatment . Electric Power
Research Institute, Palo Alto, Calif. TR-105041. April 1996.
4. Cycle Chemistry Guidelines for Fossil Plants: Phosphate Treatment For Drum Units.
Electric Power Research Institute, Palo Alto, Calif. TR-103665. Final Report,
December 1994.

6-11

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

5. Cycle Chemistry Guidelines for Fossil Plants: Oxygenated Treatment. Electric Power
Research Institute, Palo Alto, Calif. TR-102285. December 1994.
6. Sodium Hydroxide for Conditioning the Boiler Water of Drum-Type Boilers. Electric
Power Research Institute, Palo Alto, Calif. TR-104007. January 1995.
7. N. A. Styrikovich and O. I. Martynova, Contaminations of the Steam in Boiling
Reactors from Solution of Water Impurities. Soviet Atomic Energy, Vol. 15, 1963.
8. O. I. Martynova, Transport and Concentration Processes of Steam and Water
Impurities in Steam Generating Systems. In Water and Steam-Their Properties and
Current Industrial Applications. Ed. by J. Straub and K. Scheffler for International
Association for the Properties of Steam, Pergamon Press, New York, 1980.
9. Monitoring Cycle Water Chemistry In Fossil Plants: Volume 1, Monitoring Results.
Electric Power Research Institute, Palo Alto, Calif. EPRI GS-7556, Vol. 1, October
1991.
10. Monitoring Cycle Water Chemistry in Fossil Plants, Volume 3: Project Conclusions and
Recommendations. Electric Power Research Institute, Palo Alto, Calif. EPRI CS-7556,
Vol. 3, October 1991.
11. Behavior of Ammonium Salts in Steam Cycles. Electric Power Research Institute, Palo
Alto, Calif. EPRI TR-102377, Final Report, December 1993.
12. D. A. Palmer and J. M. Simonson, Volatility of Electrolytes: Application to
Water/Steam Cycles. Presented at Fourth International Conference on Fossil Plant
Cycle Chemistry. EPRI TR-104502. January 1995.
13. Assessment of the Ray Diagram. Electric Power Research Institute, Palo Alto, Calif.
TR-106017. 1996.
14. D. A. Palmer, J. M. Simonson and D. B. Joyce. Volatility of Copper. Interaction of
Non-Iron Based Materials with Water and Steam. Edited by B. Dooley and A. Bursik.
EPRI Proceedings TR-108236. July 1997.
15. G. V. Vasilenko, et al., Deposition of Copper Compounds in the Feed Circuit of
Supercritical Units, Therm. Eng. 19, 1972.
16. Z. V. Deeva, Study of the Pickup of Copper Compounds in Steam in a Supercritical
Boiler, Therm. Eng. 12, 1965.
17. T. K. Margulova and Y. M. Kostrikin, On the Mechanism of Copper Deposit
Formation, Therm. Eng. 15, 1968.
18. O. I. Martynova, Y. F. Samoilov, and T. I. Petrova, The Effect of Ammonia on the
Distribution Ratio of Copper Corrosion Products Between Water and Steam, Therm.
Eng. 16, 1969.
19. Deposition of Corrosive Salts From Steam. Electric Power Research Institute, Palo Alto,
Calif. EPRI NP-3002. April 1983.

6-12

EPRI Licensed Material


Volatility and Solubility of Copper in Steam

20. T. I. Petrova and O. I. Martynova. Behavior of Copper Corrosion Products in the


Two-Phase Region. Interaction of Non-Iron Based Materials with Water and Steam.
Edited by B. Dooley and A. Bursik. EPRI Proceedings TR-108236. July 1997.
21. M. L. Hofmann. Concerted Utility Effort Tames Copper Deposition. Power,
Volume 138, No. 6. June 1994.
22. F. J. Pocock and J. F. Stewart, The Solubility of Copper and its Oxides in
Supercritical Steam. Journal of Engineering for Power. 1963.
23. O. I. Martynova. Solubility of Inorganic Compounds in Subcritical and
Supercritical Water. in High Temperature and High Pressure Electrochemistry in
Aqueous Solutions, Surrey, England 1974. Edited by D. deG. Jones and R. W.
Staehle.
24. L. N. Varyash. Hydrolysis of Cu(II) at 25350C. Geokhimiya, p. 1003, 1985.
25. B. Hearn, M. R. Hunt and A. Hayward. Solubility of Cupric Oxide in Pure
Subcritical and Supercritical Water. J. of Chem. and Engineering Data . Vol 14, p. 442,
1969.

6-13

EPRI Licensed Material

7
COPPER DEPOSITION IN TURBINES

INTRODUCTION
Deposition of copper and its oxides on the inlet stages of high pressure turbine sections
can have significant impacts on the capacity and efficiency of fossil plant units which
employ copper alloys in the feedwater. This section addresses the routes by which
copper enters the steam path and is directed to the turbine. Characteristics of copperrich deposits which may be formed are presented, based on results of assessments by
utilities that have experienced turbine copper fouling. Theories and questions
pertaining to the mechanism of copper deposition or copper in turbines are discussed.
The problem of copper deposition was first recognized in the late 1950s with the
introduction of high pressures in supercritical units in the United States. One of the
first experiences of the deposition of copper oxides in the HP turbine was in Ohio
Power Companys Philo 6 supercritical unit in 1958(1). During the first year of operation
of this unit serious deposition occurred in the HP turbine. Within a short period of
time, the unit was shut down as a result of decreased safety factor on the thrust
pressure of the HP turbine. Heavy deposits of cupric and cuprous oxides were found
on stage wheels and diaphragms, as well as lighter deposits of the same type on the
first reheat turbine, second reheat turbine, and LP turbine. Figure 7-1 shows
distribution of deposits throughout the HP turbine. The inlet steam was at 28.9 MPa
(4200 psi). Deposition of both copper oxides occurred as the steam expanded beyond
about the third stage. This manifestation was followed shortly by a similar incident in
Cleveland Electrics Avon Unit #8 in 1960(2).
In 1971 the first unit in the former USSR to move to oxygenated treatment (OT) had
brass low pressure heaters. When the unit was converted to OT, copper levels went
from around 2 ppb to over 20 ppb at the economizer inlet and a problem subsequently
developed with copper deposition in the HP turbine.
Historically, the problem was fairly frequent in supercritical units, but has now
essentially disappeared with the replacement of copper alloys in the feedwater trains of
those units.

7-1

EPRI Licensed Material


Copper Deposition in Turbines

Source: Reference 3 from Reference 1

Figure 7-1 Solubility of Cupric Oxide at Temperatures and Pressures Corresponding to


Various Stages of the High Pressure Turbine

7-2

EPRI Licensed Material


Copper Deposition in Turbines

TRANSPORT TO THE STEAM PATH AND TURBINE


As indicated in previous sections, copper may be introduced to the steam path with
feedwater used for attemperation or as a result of carryover, both mechanical and
volatile. Although utility plant personnel should regard each source as potentially
significant until proven otherwise, the current base of scientific knowledge and field
experience suggest that volatile carryover will typically prove to be the primary
transport mechanism, particularly during normal unit service.

Importance of Individual Transport Mechanisms


To illustrate this point, it is useful to consider potential contributions of the individual
transport mechanisms to copper levels in the main steam. As an example, a range of
values for attemperation water flow, mechanical carryover, and volatile carryover have
been assumed as follows.

Attemperation water flow: 1.0, 5.0 and 10.0 percent of saturated steam flow

Mechanical carryover rate: 0.1, 0.2 and 0.3 percent of boiler water total copper
concentration (from Figure 6-1)

Volatile carryover rate: 10, 20 and 30 percent of boiler water dissolved copper
concentration (from Figure 6-2 for pressures above 2300 psi (16 MPa))

Using these assumed values, simple computations may then be performed to determine
the concentration of copper in main steam attributable to each transport path. For the
purpose of comparing potential contributions of the individual transport paths, it is
informative to fix the copper level acceptable in the steam as a constant, and calculate
the concentration of copper in feedwater (for attemperation) or boiler water (for
carryover) required to attain (but not exceed) this value.
Table 7-1 presents results of these calculations for an arbitrarily selected saturated
steam copper concentration of 2 ppb, the EPRI guideline value for copper in boiler
feedwater in mixed metallurgy units. Maximum permissible copper concentration
entries to Table 7-1 assume no contribution from the other transport mechanisms.
Naturally, this assumption is not valid for actual fossil plant units since all three
transport mechanisms will result in introduction of some copper to the steam path.
However, it does show that, under most conditions, volatile carryover is likely to
introduce most of the copper to the steam path, even when distribution ratios lower
than predicted by the ray diagram are assumed.

7-3

EPRI Licensed Material


Copper Deposition in Turbines

Table 7-1
Comparison of Available Copper Transport Paths
Transport Path

Copper Concentration*

Attemperation Water
1.0%

200

5.0%

40

10.0%

20
Mechanical Carryover

0.1%

2000

0.2%

1000

0.3%

667
Volatile Carryover

10%

20

20%

10

30%

6.7

*Maximum Permissible Copper Concentrations in Feedwater (Attemperation) or Boiler Water


(Carryover) to Maintain 2 ppb Copper in Main Steam

Under normal operating conditions with chemistry maintained in compliance with


EPRI guidelines, contribution of copper to the steam path through spray attemperation
and due to mechanical carryover would be minimal. At 10 percent attemperation and 2
ppb of copper in the feedwater the contribution to saturated steam would be 0.2 ppb.
A mechanical carryover rate of 0.3 percent with a boiler water total copper
concentration of 50 ppb (a value presently applied by at least one utility with turbine
copper fouling problems) would contribute 0.15 ppb of copper to the saturated steam.
This could obviously increase drastically if there were problems with drum level
control during startup and transients when boiler water copper levels were much
elevated. The contribution due to volatilization of copper would be more significant,
depending on both distribution ratio and the concentration of dissolved copper present
in the boiler water. The relative significance of volatile carryover is further
substantiated by utility industry experience which indicates that turbine copper fouling
generally develops at drum pressures of around 2300 2400 psig and becomes
increasingly significant at higher pressures.
Comparison of the data in Table 7-1 to concentrations of copper determined during
copper transport studies (Section 5) indicates that all transport paths may be significant
during peaking and cycling service. Limiting drum pressure at these times may be of
benefit in reducing volatile carryover. Operationally, there is often limited flexibility in
control of attemperation water flow rates if overheating of the superheater tubing is to
be avoided. Also, little can be done to minimize mechanical carryover other than to
7-4

EPRI Licensed Material


Copper Deposition in Turbines

maintain boiler steam drums internals in good condition and to exercise proper control
of drum level.
Not all copper entering the steam path is necessarily transported to the turbine. Recent
investigations have verified that deposition of copper occurs in superheaters(4, 5). In
extreme cases, copper species represent the most prevalent constituents of primary
superheater steamside solids(6). Utility plant experience at one station (Figure 7-2)
indicates that copper deposition prior to the attemperation spray injection point is
much greater than beyond the attemperation point. This observation would appear to
further corroborate the importance of volatile carryover relative to attemperation as the
primary copper transport mechanism.
Copper deposits in the primary superheater could, potentially, reenter the steam and be
directed to the turbine. Accounts of copper transport in several fossil plant units
indicate that copper deposits in feedwater heaters and boiler waterwalls reenter the
feedwater and boiler water during unit startups (Section 5) and it would appear that a
similar phenomenon could occur in the steam path at locations subject to deposition,
such as the primary superheater.
Release of copper from the superheater could be either chemical (by volatilization,
especially under oxidizing conditions) or physical (via solids exfoliation due to metal
temperature variations) in nature. Both phenomena could be significant.
Determination of the relative significance of each would require sampling and analysis
of copper in saturated and main steam samples under transient conditions. Use of wet
layup procedures or chemical treatment of water used for hydrostatic testing also may
influence release of previously deposited copper from superheater tubing.

7-5

EPRI Licensed Material


Copper Deposition in Turbines

Source: Reference 6

Figure 7-2 Distribution of Copper in a Fossil Unit


DEPOSITION IN TURBINES
Turbine copper deposits are a significant cause of turbine performance losses(710).
Deposit accumulations affect both clearances and surface smoothness (Figure 2-7)
producing losses in efficiency and capacity. The staggering effects of copper deposits
have been delineated in Section 2. Other factors likely to degrade turbine performance
include physical damage, erosion, and steam leakage. Steam path audits, conducted
during turbine overhauls, provide a basis for quantitatively designating losses as either
7-6

EPRI Licensed Material


Copper Deposition in Turbines

recoverable or unrecoverable and determining the contribution of the various causes of


performance deterioration(11). Utilities should consider it good practice to conduct
steam path audits during scheduled turbine overhauls.

Effect of Deposits on Capacity


In many instances, detrimental effects of turbine deposits are limited to efficiency
losses(7). However, when copper transport activity is significant, capacity losses have
also been experienced(5, 1216).
One example occurred on a drum unit with admiralty brass condenser tubing and LP
heaters tubed with Monel(17). The unit was operating with conventional deoxygenated
AVT. The hydrazine feed was stopped. For approximately four months there was no
indication of a problem as determined by the measurement of copper levels at the
economizer inlet. For another couple of months, scatter in the measurements made it
difficult to determine whether the rising copper levels were real. Within another two
months, however, the copper levels at the economizer started to rise above the 2 ppb
limit. At approximately the same time, the saturated steam copper levels rose
dramatically above the historical, undetectable levels to over 11 ppb (Figure 7-3). The
unit started to lose MW capacity (Figure 7-4). The hydrazine was turned back on at
very high levels, with the result that the copper levels in both the feedwater and
saturated steam started to return to the normal low levels (Figure 7-3). However, the
MW capacity continued to drop (Figure 7-4), and eventually the HP turbine had to be
chemically cleaned.

Figure 7-3 The effect of the elimination of N2H4 from the feedwater of a mixedmetallurgy feedwater system on the copper transport in feedwater and saturated steam.
Adapted from reference 17.
7-7

EPRI Licensed Material


Copper Deposition in Turbines

Figure 7-4 Decrease in gross MW capacity after eliminating the oxygen scavenger N2H4
from the feedwater. Note that the MW keeps decreasing even after the N2H4 has been
restored (A). Adapted from reference 17.
High pressure turbine deposits reduce nozzle area and flow capacity; this relationship
is shown in Figure 7-5(9). Using the figure, it may be determined that first stage
deposits reducing nozzle area by ten percent would reduce the capacity of an impulsetype turbine by about three percent. Effects of deposits on more than one stage are
approximately cumulative in the absence of other factors which influence performance.

Turbine Copper Deposits and Steam Path Audits


Conducting steam path audits provide an excellent means to quantify the cost penalties
associated with turbine copper deposits. The findings present utility management
personnel with a clear picture of the savings possible if deposition activity can
effectively be controlled. The experience in Units 1 and 2 at Craig Station of Tri-State
Generation and Transmission represents an excellent case history of how steam path
audit findings were used to justify improvements which were reported to offer a net
payback acceptable to utility management(13, 14).
The annualized cost of efficiency losses due to turbine deposits was reported to be in
excess of $230,000 in Unit 1 and in excess of $630,000 in Unit 2. Most of these costs
were associated with efficiency losses in the high pressure turbine. Copper was the
most significant turbine deposit constituent, accounting for 71% and 46% of the high
pressure section capacity losses in Units 1 and 2, respectively. Evaluation of these
findings by station personnel responsible for performance and chemistry resulted in the
following action plan.

7-8

Conversion of Units 1 and 2 from phosphate to all-volatile boiler water treatment to


eliminate mechanical carryover of sodium phosphate.

EPRI Licensed Material


Copper Deposition in Turbines

Installation of a shared condensate polisher (powdered resin type) for use in


startups of Units 1 and 2 to remove copper and other feedwater corrosion products
as well as sulfate and silica which would otherwise be transported to the boiler and
turbine.

Figure 7-5 Effect of Nozzle Area Change on Flow Capacity of Impulse-Type


Turbines(9).
(This figure is also applicable to the first stage of reaction-type turbines)
Favorable operating experience of Unit 3 at Craig Station - which includes steel heaters
and a powdered resin unit - strongly supported the second part of the action plan.
Assessment of available condensate polisher options determined that a retrofit system
could be economically justified if the action plan was effective in reducing turbine
deposition by 75%. A suitable polisher system was designated, designed, constructed
and placed in service approximately four years after the initial steam path audit.
Station personnel indicated that performance and efficiency losses quantified by the
steam path audits provided most of the financial incentive favoring the polisher system
retrofit. Other benefits which were expected included reduced need for replacement
power, reduced chemical cleaning costs and reduction or elimination of other
miscellaneous costs.

Deposit Characteristics
A survey of turbine deposits was conducted in the early 1980s as part of an effort to
assess failures in the next to last row (L-1) of turbine blades in reheat fossil units(18).
This effort involved analysis of more than 800 deposits from over 100 turbines.
Analysis of the turbine deposits identified in excess of 120 chemical species; deposit
constituents which included copper are indicated in Table 7-2.

7-9

EPRI Licensed Material


Copper Deposition in Turbines

Table 7-2
Copper Species Identified by a Turbine Deposit Survey
Formula

Name

Formula

Name

CuO

Cupric Oxide

CuAlO2

Copper Aluminum Oxide

Cu2O

Cuprous Oxide

CuSO4(OH)6

Copper Sulfate Hydrate

Cu

Metallic Copper

CuFe(OH)2SO4

Copper Iron Sulfate Hydrate

FeCu

Metallic

Na2Cu(SO4)2

Sodium Copper Sulfate

Cu6Fe3O7

Copper Iron Oxide

CuCO3Cu(OH)2

Copper Carbonate, Basic

FeCuO2

Copper Iron (II) Oxide

Cu2(OH)3Cl

Copper (II) Oxychloride

Source: Reference 18

Results for 260 deposits from 41 turbines at drum boiler units utilizing phosphate
chemistry revealed copper oxides to be the most commonly encountered copper
species; metallic copper was occasionally reported (Table 7-3).
Table 7-3
Forms of Copper Reported in Turbine Deposits from Drum Boiler Units
Compound

Frequency of Occurence, Percent


Deposits with
Compound

Units with
Compound

CuO

28.5

34.1

Cu2O

21.9

19.5

FeCuO2

3.5

9.8

Cu

1.5

2.4

Source: Reference 18

The copper species were present primarily in high pressure turbine deposits. To a
lesser extent they were identified in the intermediate turbine (cupric and cuprous
oxides) and in a few cases in low pressure turbine deposits (metallic copper, cupric
oxide and cuprous oxide). Summary data for the drum boiler units addressed by this
survey indicate that the average copper concentration in turbine deposits was 11.5
percent with a maximum value of 96.9 percent(18).
The stoichiometries of all these copper compounds in Tables 7-2 and 7-3 provide no
information on the composition of the copper species that transported in the steam. But
they do indicate the role of complexing anions in enhancing the solubility of copper in
the boiler water and feedwater as discussed in Section 6.

7-10

EPRI Licensed Material


Copper Deposition in Turbines

Published accounts of copper-rich high pressure turbine deposits in units with drum
boilers typically indicate the copper to be present predominantly as cupric oxide (CuO).
At some stations, the presence of copper as cuprous oxide (Cu2O) in turbine deposits is
also indicated. In some cases, it appears an attempt was made to ascertain the actual
species present while in others the indicated presence of copper in turbine deposits as
cupric oxide may have been reported by convention.
Tables 7-4 to 7-7 are published results of analyses for turbine deposits from units which
have experienced capacity losses as a result of copper fouling. Table 7-4 presents
deposit analysis results for various stages of the Unit No. 4 high and intermediate
pressure turbine of Long Island Light Companys Northport Station(12). Table 7-5
includes similar data for Units Nos. 1 and 2 of the Columbia Generating Station of
Wisconsin Power and Light Company(5). Composition data for turbine deposits in the
Units 7 and 8 turbines of Cincinnati Gas and Electric Companys Miami Fort Station are
shown in Table 7-6(15). Turbine deposits analyses conducted as part of steam path
audits conducted on Unit Nos. 1 and 2 of Tri-State Generation and Transmissions
Craig Station are given in Table 7-7(13).
Analysis of these and other turbine deposit characteristics raises the following points:

Copper oxides (frequently reported as CuO) are the primary constituents of high
pressure deposits; they may be present as minor constituents of intermediate
turbine deposits.

Appreciable amounts of sodium phosphate compounds can be present in high


pressure turbine deposits if the boiler employs phosphate chemistry.

Composition of corresponding rotating and stationary blade deposits is generally


similar.

Table 7-4
Turbine Deposit Composition: Northport Station, Unit No. 4
Sample Source

Composition, Weight Percent As


CuO

Na3PO4 Na2SiO3
High Pressure Turbine

NiO

Fe3O4

Upper Nozzle Box Vanes

84

7.3

6.4

--

2nd Stage Diaphragms

81

6.1

1.9

4.4

--

6th Stage Diaphragms

85

4.2

2.7

--

Intermediate Pressure Turbine


8th Stage Bucket Covers

2.9

2.4

--

0.9

95

10th Stage Buckets

2.2

3.8

93

1.3

--

11th Stage Buckets

2.8

4.1

95

1.2

--

12th Stage Buckets

3.2

4.1

96

1.3

--

Source: Reference 12

7-11

EPRI Licensed Material


Copper Deposition in Turbines

Table 7-5
Turbine Deposit Composition: Columbia Generating Station, Units Nos. 1 and 2
High Pressure Turbine

Composition, Weight Percent As


Cu

Fe

Na

UNIT NO. 1 - Sample Source


Nozzle Box

>96

<1

<1

<1

1st Stage Blades

>96

<1

<1

<1

2nd Stage Blades

>90

<1

<1

3rd Stage Blades

>90

<1

<1

4th Stage Blades

>90

<1

5th Stage Blades

>94

<1

<1

<1

6th Stage Blades

>96

<1

<1

<1

CuO

Fe2O3

Na2O

P2O5

7th Stage Covers

61

32

<1

8th Stage Covers

61

30

<1

9th Stage Covers

55

37

<1

10th Stage Covers

54

38

<1

11th Stage Covers

59

33

<1

Intermediate Pressure
Turbine

UNIT NO. 2 - Sample Source


High Pressure Turbine

Composition, Weight Percent As


CuO

Fe2O3

Na2O

P2O5

Nozzle Box

94

<1

2nd Stage Blades

58

22

16

2nd Stage Diaphragms

88

2nd Stage Blade Covers

22

39

31

3rd Stage Blades

31

38

26

4th Stage Blades

34

37

25

5th Stage Blades

31

40

24

6th Stage Blades

26

44

25

CuO

Fe2O3

Na2O

P2O5

82

<1

Intermediate Pressure
Turbine

7th, 8th, 9th Stage Blades

Source: Reference 5

7-12

EPRI Licensed Material


Copper Deposition in Turbines

Table 7-6
Turbine Deposit Composition: Miami Fort Station, Units Nos. 7 and 8
Sample Source

Composition, Weight Percent As


CuO

Fe3O4

Other

UNIT NO. 7 (1986)


High Pressure Turbine
Second Stage Rotor/Diaphragm

77/75

1/1

22/24

Third Stage Rotor/Diaphragm

68/77

2/1

30/22

Fourth Stage Rotor/Diaphragm

63/71

2/1

35/28

Fifth Stage Rotor/Diaphragm

61/77

2/1

37/22

Sixth Stage Rotor/Diaphragm

62/76

3/1

36/23

--

--

--

Third Stage Rotor/Diaphragm

87/88

4/4

9/8

Fourth Stage Rotor/Diaphragm

84/87

3/3

13/10

Fifth Stage Rotor/Diaphragm

76/84

4/3

20/13

Sixth Stage Rotor/Diaphragm

71/81

5/3

24/16

Seventh Stage Diaphragm

94

Eighth Stage Diaphragm

89

Ninth Stage Diaphragm

87

Tenth Stage Diaphragm

88

Eleventh Stage Diaphragm

88

UNIT NO. 8 (1987)


High Pressure Turbine
Second Stage Rotor/Diaphragm

Intermediate Pressure Turbine

Source: Reference 15

7-13

EPRI Licensed Material


Copper Deposition in Turbines

Table 7-7
Turbine Deposit Composition: Craig Station, Units Nos. 1 and 2
Sample Source

Composition, Weight Percent As


CuO

Fe2O3

NaO P2O5 SO4

SiO2

Misc.

UNIT NO. 1 (1991)


High Pressure Turbine
Summary Results

71

--

--

10

85

--

--

--

--

--

--

--

--

--

86

14

15

--

--

Intermediate Pressure Turbine


Summary Results
Low Pressure Turbine
Summary Results

UNIT NO. 2 (1989)


High Pressure Turbine
Summary Results

46

27

Intermediate Pressure Turbine


Summary Results

No significant deposition observed

Low Pressure Turbine


Summary Results

--

--

--

--

--

87

13

Source: Reference 13

The presence of sodium phosphate species in the high pressure turbine of units with
phosphate treated boilers is an indication of mechanical carryover activity. Evaluation
of steam drum separator integrity and performance, review of drum level control and,
potentially, optimization of boiler water treatment should be implemented to reduce
mechanical carryover and deposition in the turbine; copper deposition also may be
reduced, but only to the extent that correction of any mechanical deficiencies allows.

Mechanism of Deposition
The mechanism by which copper deposits are formed within turbines and other
components of fossil plant units is not fully understood. With respect to steam path
deposition, the form in which copper exists (cupric oxide, cuprous oxide, metallic
copper) and the solubility of the individual species in steam appear to be significant
influencing factors. The investigation of steam solubility conducted by Pocock and
Stewart(3) discussed in Section 6 (Figure 6-4) determined tenorite (CuO) to be more
soluble than either cuprite (Cu2O) or metallic copper. Figure 7-6 indicates the predicted
maximum apparent solubility of tenorite as a function of specific volume, based on data
collected during experiments conducted at subcritical and supercritical conditions. The
relationship between cupric oxide solubility and pressure is shown in Figure 6-4 for pH
7-14

EPRI Licensed Material


Copper Deposition in Turbines

of 9.5. The solubility is enhanced at the higher pH which was established by treatment
with ammonia.

Figure 7-6 Solubility of Cupric Oxide in Steam as a Function of Specific Volume(3)

The solubility of copper in steam decreases as a result of pressure reduction and/or


increases in specific volume; this is exactly what occurs as steam is expanded in the
high pressure turbine. It would not, however, appear to account for the extensive
copper deposits identified in the primary superheater section of some units which have
experienced turbine copper fouling. The influence of steam temperature on copper
solubility (at constant pressure) does not appear to be significant(3). It would therefore
appear that other factors influence deposition of copper in superheaters.
Observations made by Jonas(4) in comparing the copper content of primary superheater
and turbine deposits in a utility unit which had experienced significant capacity losses
due to turbine fouling include the following.

Chemical compounds found in the superheater deposits included Fe3O4, Fe2O3, FeO,
Fe, FeZn, Cu and Cu2O.

Compounds found in high pressure turbine deposits included Fe3O4, Fe2O3, Fe,
FeCuO2, Cu, Cu2O, CuO, Na3PO4, NaOH, Al2O3 and SiO2.
7-15

EPRI Licensed Material


Copper Deposition in Turbines

The fact that tenorite (CuO) was identified present in the turbine deposits but not in the
superheater deposits may simply reflect the higher solubility in steam determined for
this species. However, it should be kept in mind that the reduced forms of copper
(metallic and cuprite) are also believed to be less volatile than tenorite. The sodium
salts identified in the turbine deposits are even less volatile and typically enter the
steam path as a result of mechanical carryover. Reduced forms of copper, by virtue of
their low volatility and solubility relative to tenorite, may enter the superheater by
mechanical carryover. Transport would be potentially significant during startups.
Adding superheat to the saturated steam would result in evaporation of the liquid;
total copper in the superheated steam may, therefore, be present at levels exceeding the
maximum soluble concentration. Reduction of tenorite to cuprite within the
superheater could also account for the deposit characteristics reported. Corrosion of
steel (oxidation of iron) to release hydrogen would appear to be the coupled oxidation
reaction.
Copper species not dissolved in steam are present primarily as fine particulates. These
particles, including some of which would be retained on a 0.45 micron fillers have been
reported to behave as colloids(19). Deposition phenomena, including that of copper in
turbines, was investigated(19) to determine if the mechanism involved electrokinetic
effects, wherein opposite electrostatic charges existed on the colloidal particles in the
steam path and oxides on the surface of the turbine. For a detailed account of transport
properties and electrokinetic effects, readers should consult Section 7-4 of Reference 20.
More extensive evaluations of deposition have been conducted for species present in
aqueous solutions. The double layer theory, originally published in 1924, indicates that
some mobile ions in solution will be attracted to a surface and become attached while
others remain mobile with respect to the surface. The inner layer is naturally compact
while the outer layer is diffuse. The difference in potential of the shear phase
(boundary between the compact and the diffuse layers) and the bulk solution is termed
the zeta potential. Measurement and control of zeta potential and streaming current
are useful in evaluation of deposition and dispersion phenomena. Factors known to
influence interaction of charged particles and surfaces in aqueous solutions include
conductivity, temperature and fluid velocity.
The zero point of charge (ZPC) is defined as the pH at which surface charge becomes
neutral. It may be related to the isoelectric point of surface (IEPS), that is the pH at
which the surface of the compound has an equal tendency to release positive and
negative ions. Table 7-8 provides IEPS values determined for several compounds in
neutral water. IEPS values above 7.0 are indicative of a positive surface charge while
values below 7.0 indicate a negative surface charge would exit. The table suggests that
cupric oxide particles would be positively charged while the charge of iron species
varies with the form actually present. Magnetite particles exhibit a slightly negative
IEPS, however, the IEPS of an extended metal surface (such as turbine stage metal) is
more difficult to predict. The presence or absence of an oxide film would influence the
7-16

EPRI Licensed Material


Copper Deposition in Turbines

tendency of the surface to exhibit cathodic or anodic behavior; galvanic complexing to


other alloys could also influence the situation.
Table 7-8
Isoelectric Point of Surface (IEPS) Values for Several Compounds
Compounds

IEPS

CuO

9.5 0.4

NiO

10.3 0.4

Ni(OH)2

approx. 11-12

FeO OH

5.4 - 7.3

Fe2O3

6-8

Fe(OH)2

12 0.5

Fe3O4

6.5 0.2

Cu(OH)2

11.40

Source: Reference 21 as reported in Reference 19

As a result of investigations by Gasparini, Della Rocca and Ioannilli(19), the following


conclusions were reached.

Particles physically transported to the steam (presumably by either attemperation


water or mechanical carryover) would exist as charged bodies.

Molecules of metal oxides dissolved in steam would be subject to precipitation as a


result of expansion in the turbine; this material could coalesce into aggregates which
would form deposits on nearby surfaces.

The existence of charged particles in superheated steam-condensate was verified


experimentally.

CONCLUDING REMARKS ON COPPER DEPOSITION IN TURBINES


It would appear that deposition of copper in turbines could be influenced by certain
design factors such as materials choices, fluid velocity and the existence of galvanic
couples. Radical change in turbine design to reduce copper fouling tendencies is
probably impractical, however certain modifications of existing designs might be
beneficial. Development of suitable coatings for application to turbine surfaces
(thereby adjusting the charge and tendency to attract copper) may be one approach
which could eventually become technically feasible and economically justifiable to
utilities. Adjustment of steam chemistry might also influence deposition tendencies but
collective utility experiences would seem to suggest that turbine deposition can be
experienced under several treatment regimes. However, a better understanding of the
mechanism by which copper deposits in turbines is urgently needed to establish criteria
7-17

EPRI Licensed Material


Copper Deposition in Turbines

which may be used to develop and evaluate potential corrective measures intended to
reduce or eliminate the deposition activity.
REFERENCES
1. T. T. Frankenberg, A. G. Lloyd, and E. B. Morris. The Second Year of Operating
Experience with the Philo Supercritical Operating Pressure Unit. Proceedings of the
American Power Conference, Volume 23. 1959.
2. N. F. Gill and N. D. Flack. Initial Operation of Avon #8, a Supercritical Plant.
Proceedings of the American Power Conference, Volume 25. 1961.
3. F. J. Pocock and J. F. Stewart, The Solubility of Copper and its Oxides in
Supercritical Steam. Journal of Engineering for Power, 1963.
4. O. Jonas, L. Dupee, T. Gilchrist, G. Lawrence and M. Stevens. Copper Deposition
and MW Loss: Problems and Its Solutions Proceedings: 57th Annual Meeting,
International Water Conference. Engineers Society of Western Pennsylvania, 1996.
5. G. S. Lawrence. Chemical Cleaning of HP Turbines at Columbia Energy Center.
Proceedings: 56th Annual Meeting, International Water Conference. Engineers Society of
Western Pennsylvania. 1995.
6. G. S. Lawrence. Copper Fouling of Turbines at Columbia Generating Station.
Presented at the 1996 Meeting on Turbine Copper Fouling in Salt Lake City, Utah.
(Unpublished).
7. R. E. Brandon. Diagnosis of Turbine Problems as Determined from Monitoring
Results. Presented at the Heat Rate Improvement in Existing Plants Workshop .
September 20-22, 1983.
8. K. C. Cotton. Evaluating and Improving Steam Turbine Performance . Cotton Fact Inc.,
1993.
9. P. Schofield. Efficiency Maintenance of Large Steam Turbines. Presented at the
Pacific Coast Electrical Association 1982 Engineering and Operating Conference .
10. Forster, V. T. Performance Loss of Modern Steam-Turbine Plant Due to Surface
Roughness. Proc. Inst. Mech. Engrs., 1966-67, Vol. 181, Pt. 1, No. 17.
11. Cycle Chemistry Improvement Program. Electric Power Research Institute, Palo Alto,
Calif. TR-106371. April 1997.
12. B. T. Hagewood. CuO Fouling Northport 4 High-Pressure Steam Turbine. Second
Conference on Fossil Plant Cycle Chemistry. Electric Power Research Institute, Palo
Alto Calif. August 1988.
13. R. Devalois, T. Gilchrist and K. Price. Justification of the Retrofit of a Condensate
Polishing System at Tri-State G&Ts Craig Station. Presented at EPRI Condensate
Polishing Workshop: Deep Bed and Powdered Resin Systems. New Orleans, LA,
September 15-17, 1993.
7-18

EPRI Licensed Material


Copper Deposition in Turbines

14. T. Gilchrist. Preliminary Steam Path Audit and Turbine Deposition Analysis
Results at Tri-State G&Ts Craig Station. Presented at the Fourth Annual Southwest
Chemistry Workshop, Farmington, New Mexico, July 26-27, 1995.
15. M. L. Hofmann. Copper Transport at Miami Fort Station Fourteenth Annual
Electric Utility Workshop. Champaign, Illinois, 1994.
16. S. G. Sawochka, M. E. Clouse, B. Larson and T. Archbold. Corrosion Product
Control at the Laramie River Station. Proceedings: 53rd Annual Meeting, International
Water Conference. Engineers Society of Western Pennsylvania 1992.
17. T. Gilchrist. Hydrazine Feed and its Impact on Corrosion Product Transport at
Craig Station. Fifth Annual Southwest Chemistry Workshop , St. Johns, Arizona. July
1996.
18. Characterization of Operational Environment for Steam Turbine-Blading Alloys. Electric
Power Research Institute, Palo Alto, Calif. CS-2931. August 1984.
19. R. Gasparini, C. Della Rocca and E. Ioannilli. A New Approach to the Study and
Prevention of Deposits in Modern Power Stations. Combustion, November 1969.
20. The ASME Handbook on Water Technology for Thermal Power System : Electric Power
Research Institute, Palo Alto, Calif. EPRI CS-6303. 1989.
21. G. A. Parks. Chem. Rev., Vol. 65, 1965.

7-19

EPRI Licensed Material

8
SOLUTIONS FOR COPPER PROBLEMS

INTRODUCTION
The previous sections have provided information on copper around the cycle:

Corrosion of copper alloys in feedwater (Section 3)

Electrochemistry of copper alloys in feedwater (Section 4)

Transport of copper oxides in feedwater (Section 5)

Volatility and solubility of copper and its oxides in steam (Section 6)

Copper deposition in turbines (Section 7)

Section 9 brings together all the sections to provide a complete assessment and effective
corrective actions of the problems of copper release, transport and deposition. It
provides a generic sequential approach (road-map) for utilities to follow.
This section is dedicated to delineating the various possible solutions to the problem.
Table 8-1 breaks the solutions into three groups.
Clearly the only way to completely eliminate copper problems is to change the
feedwater heaters to all-ferrous materials (Solution 8). There are a number of excellent
examples where utilities have justified changing heaters(1,2). The opposite extreme is to
adopt only the palliative approaches under Deposit Removal (Solutions 13). If new
heaters cannot be cost justified, then the overall approach must involve optimizing the
feedwater chemistry to minimize the copper corrosion and transport together with
either/or/and deposit removal (Solutions 13) and equipment changes (Solutions 8
11). Although cycle design changes are included in Table 8-1, utilities clearly have not
been able to justify piping changes (Solution 9) or the installation of startup loops
(Solution 10). An increasing number of utilities(1,3) have however been able to justify
the installation of condensate polishers (Solution 11).

8-1

EPRI Licensed Material


Solutions for Copper Problems

Table 8-1
Possible Solutions for Copper Problems
A. Deposit Removal (short-term actions)
Solution 1. Cleaning turbine
Solution 2. Cleaning feedwater heaters
Solution 3. Cleaning boiler
B. Chemistry Approaches
Solution 4. Optimize feedwater chemistry
Solution 5. Reduce/control air-inleakage
Solution 6. Maintain reducing feedwater environment (ORP < 0 mV)
Solution 7. Develop layup/shutdown/startup procedures
C. Equipment Changes
Solution 8. Change feedwater heater materials
Solution 9. Change cycle design/piping changes
Solution 10. Install startup loop or turbine by-pass
Solution 11. Install condensate polisher

A. DEPOSIT REMOVAL (SHORT TERM SOLUTIONS)


The removal of copper deposits from the equipment should be regarded as a temporary
fix to the overall problem. A very extensive literature base is available on chemically
cleaning of boiler waterwalls. The reader is referenced to these materials(4,14). Tables 82 and 8-3 provide information on typical solvents used and their capacity for removing
a wide range of deposits. It should be noted that table 8-2 includes hydroxyacetic
formic acid and hydrofluoric acid for completeness; this does not imply these have the
ability to remove copper as indicated in Table 8-3.

8-2

EPRI Licensed Material


Solutions for Copper Problems

Table 8-2
Typical Concentration and Temperature of Use for Common Solvents Used for Waterwall and
Economizer Cleanings
Hydrochloric

Hydroxyacetic

Ammoniated

Ammoniated

Ammonium

Hydrofluoric

Acid

Formic Acid

EDTA

Citric Acid

Bromate

Acid

Typical
Concentration

5-6%

3%

36%

3%

0.10.5%

12%

Typical
Temperature,
C (F)

6671C
(150160F)

93C
(200F)

For Iron:
8293C
(180200F)
or
(129149C
(265300F)

For Iron:
8293C
(180200F)
or
(116135C
(240275F)

For Copper:
66C (150F)

For Copper:
66C (150F)

Table 8-3
Capacity of Selected Solvents for Constituents in Waterwall and Economizer Deposits
Constituent in

Hydrochloric

Hydroxyacetic

Ammoniated

Ammoniated

Ammonium

Hydrofluoric

Deposit

Acid

Formic Acid

EDTA

Citric Acid

Bromate

Acid

Iron Oxides

High

High

High

High

N/A

High

Copper
Metallic
Oxide

Lowa
Mediuma

Traceb
Traceb

Medium
Medium

Medium
Medium

High
High

Low (for CuO)

Nickel Oxides

High

High

High

Trace

Zinc Oxide

High

High

High

Trace

N/A

Calcium Salts

High

Low

Low

N/A

N/A

Magnesium
Salts

High

Low

Low

N/A

N/A

Silica

Low

N/A

N/A

Organics

Trace

N/A

N/A

aWith

sufficient copper complexor present, such as thiourea.


copper will redeposit on bare steel surfaces.

bDissolved

8-3

EPRI Licensed Material


Solutions for Copper Problems

The reader is also referenced to the EPRI Chemical Cleaning Guidelines for information
on removing copper and its oxides from all-ferrous and copper-based feedwater heater
tubes(4).
Discussion on removing deposits from HP blade surfaces is covered in the next
subsection.

Solution 1. Cleaning Turbine


There are essentially three main methods of removing deposits from the HP turbine:
grit blasting, foam chemical cleaning, and chemical cleaning. The first two are
extensively covered in EPRIs Chemical Cleaning Guidelines(4).

Grit Blasting
The average interval between major turbine maintenance outages is 56 years. These
outages require disassembly of the turbine for inspection and maintenance. If deposits
are apparent on the blades these can be effectively removed by grit blasting, with either
aluminum oxide or other suitable material. Grit-blasting produces a soft, grey, satin
finish and slightly increases the fatigue strength of the material. Mild sand blasting has
been used by at least one utility routinely to remove copper deposits(7). Provided no
loss in turbine efficiency due to deposition is detected during unit operation, this
practice of inspecting and grit-blasting every five to six years will suffice and is
preferred to chemically cleaning the turbine.

Water Washing to Remove Soluble Deposits


For water-soluble, noncorrosive turbine deposits such as sodium phosphate, waterwashing can restore lost efficiency. Although the focus of this report is copper and its
oxides, there is a number of recent experiences, that water washing has restored some
MW loss and efficiency related to copper deposition. Actually, there are two very
different procedures which have been called water-washing. The first involves injecting
high purity water in the steam upstream of the turbine as the unit is shutting down in
order to scour deposits from turbine blading. While this steam/water-washing is
occasionally performed, the potential for vibration damage and deposition of material
in the reheater and intermediate pressure turbine makes this an undesirable approach.
For large, modern turbines, steam/water-washing should be undertaken only with
great caution and under special circumstances(5,6). Another water-washing procedure
for removing noncorrosive, water-soluble deposits involves allowing the turbine to cool
and then filling it up to the turbine shaft (through a shell drain) with high purity water
while the turbine is on turning gear. The size of the turning gear motor should be
8-4

EPRI Licensed Material


Solutions for Copper Problems

checked so that the gear is not overloaded by the windage of the blades in the water. A
temperature difference in the water may assist in removing the deposit, but the
temperature differential (or the cooling rate) should not exceed turbine manufacturer
recommendations. After concentrations have leveled, the water is drained out and the
process may be repeated until pickup is minimal.
A modified version of this procedure involves continuous recirculation of
demineralized water from the condensate pumps (e.g. 76 L/m [20 gpm] of water for a
turbine with a normal steam flow of 1.8 x 106 Kg/hr [4x106 lbs/hr] into the nozzle
block. Water can be returned to the condenser via a drain line on the cold reheat line.
This water recirculation can be continued for a twelve hour period. The system should
then be drained and refilled and recirculation repeated.
Water-soluble deposits in turbines also have been removed with foamed water rinses at
8288C (180190F). The procedure for cleaning with foamed water is essentially the
foam cleaning procedure used for copper removal minus the ammonia-based
compounds or oxidants (see next sub-section).

Chemical Cleaning
HP turbines have been successfully cleaned to remove copper deposits for many years(813). Chemical cleanings are employed to clean the turbine during periods between
turbine maintenance outages. The advantage of chemical cleaning over grit-blasting in
this case is that chemical cleaning does not require disassembly of the unit. Primarily,
the need to chemically clean would be based on losses in turbine efficiency or capacity
due to deposition on the turbine. It should be stressed that a turbine chemical cleaning
should be viewed only as a temporary solution to regain efficiency or MW capacity. If
performed in isolation without addressing the other areas such as superheater, boiler,
turbine and the operating and shutdown chemical environments, then the problem of
deposition in the turbine will return.
Solvent selection for HP turbines with deposits can be difficult because collecting
samples usually necessitates disassembling the turbine.
The standard method of chemical cleaning uses a foam cleaning operation. Choice of
solvent is critical and will depend on the composition of the deposit and the
compatibility of the solvent with turbine components. A typical solvent consists of
about 56% ammonium hydroxide, 1216% ammonium bicarbonate, and sufficient
foaming agent(4). Proprietary solvents for copper and copper oxide removal are also
available. The addition of hydrogen peroxide to the foamed mixture described above
is used when copper base metal is present. The hydrogen peroxide converts the copper
to copper oxide and enables dissolution by the ammonia.

8-5

EPRI Licensed Material


Solutions for Copper Problems

A typical cleaning procedure for removing copper oxides and copper from an HP
turbine using a foamed solution of ammonium hydroxide and ammonium bicarbonate
(and possibly an oxidant if copper metal is present) includes the following events(4):

The turbine system is on turning gear and is heated with steam or cooled with air to
the desired metal temperature. Metal temperatures should be monitored in the
impulse chamber or first stage of the turbine. All metal temperatures should be
maintained within 6577C (~150170F). All equipment and piping necessary for
the cleaning are connected and in safe, working order, and the turbine cleaning
circuit is completely isolated.

Water and foaming agents are foamed with air and injected at 65C (150F) into the
turbine to ensure leakage is not occurring through drain lines, etc. Each governor
valve is opened for 15 minutes, and antifoam is injected to collapse the spent
precleaning foamed rinse.

Aqueous ammonia, ammonium bicarbonate and foaming agent are blended and
foamed with air and injected at 65C (150F) into the turbine. Foamed solvent is
quickly and separately routed through each part of the high pressure turbine by
opening each governor valve for one hour. Operating these valves and keeping the
turbine on turning gear help ensure solvent contact with all surfaces in the high
pressure turbine. Concentrations of copper and solvent strength are monitored in
the spent, collapsed solvent during the copper removal stage. This is continued
until copper concentrations level out at a relatively low value. This stage typically
lasts about four hours.

If copper metal is present, hydrogen peroxide may be blended into the foam during
the solvent foaming step and injected into the turbine. Great care should be
exercised when handling hydrogen peroxide. Temperatures at the point of
hydrogen peroxide injection into the liquid should not exceed 65C (150F). The
copper concentration is monitored in the spent, collapsed solvent until copper
removal is complete.

The turbine is rinsed. An initial foam rinse is recommended to remove the bulk of
the material and to minimize water usage and wastewater production. The turbine
is flushed with high purity water until effluent conductivity is below the
predetermined value (approximately 20 S/cm) and no suspended solids are
apparent. Governor valves should be operated to ensure flow through all areas of
the turbine. It is recommended that at least one steam flush be performed to
volatilize any ammonia trapped in crevices in the turbine. Alternating steam
flushes with water at 3060 minutes intervals have also been used.

All lines leading to other equipment or instruments should be drained and flushed
with high purity water. Lines that cannot be flushed with water should be purged
with air or steam.

Safety is paramount throughout the planning and cleaning process. Ammonia


vapors can be toxic and explosive, contact of aqua ammonia with the skin and eyes
can cause chemical burns if not washed off promptly. Material Safety Data Sheets

8-6

EPRI Licensed Material


Solutions for Copper Problems

(MSDS) for chemicals utilized should be reviewed. Copies of MSDS for chemicals
and safety equipment should be available in a centralized and protected location in
the event of an emergency. The turbine bay must be well ventilated during the
cleaning. Running hoses through open windows on the turbine deck is an
unacceptable means of venting ammonia fumes. Piping fumes into a vacuum
system which vents gases out the top of the roof of the boiler building is
recommended. It is suggested that a trial run of injecting just foamed water be
performed to determine if there are any areas of leakage prior to the start of the
cleaning. Detailed written procedures should be prepared for the cleaning, which
include a safety meeting to discuss safety concerns. Following the cleaning, all lines
and areas in which welding will be performed must be thoroughly purged with air
to minimize explosion hazards.
Chemical cleaning of copper deposits has also been conducted without the use of
foaming agents(7). Here, the turbine rotor was slowly turned (1 to 2 rpm) and the
solution drained out of the shaft seals at both ends. The turbine was filled with
demineralized water, ammonia was added until the pH reached 10.2. Then hydrogen
peroxide was added, and after a few minutes the complexing agent EDTA was added.
The chemical concentrations in the final solutions were 1520 mg/L ammonia, 0.1%
hydrogen peroxide, and 3 g/L of EDTA. The uptake of copper (and iron) in the
solution was followed for 67 hours until the uptake leveled off. The turbine was
rinsed with demineralized water and drained several times.
B. CHEMISTRY APPROACHES

Solution 4. Optimize Feedwater Chemistry for Mixed-Metallurgy Systems


As discussed in Section 2, the selection and optimization of the feedwater chemistry is
vital to the reduction and elimination of copper problems in the cycle. Table 2-6
provides the chemistry guideline limits at the economizer inlet for mixed-metallurgy
systems. These, in conjunction with the Guiding Principles also provided in Section 2,
indicate that every effort should be given to achieving less than 2 ppb of copper during
normal operation. This subsection provides an outline of how to achieve this; for more
detailed further information the reader is referenced to the EPRI Selection and
Optimization document(15). It should be noted that optimization of the operating
feedwater will not, by itself, solve all the copper problems in the cycle; it is also
necessary to address the maintenance of a reducing environment during shutdown and
layup.
The feedwater treatment dilemma for mixed metallurgy feedwater cycles is that the
optimum pH for corrosion control of copper alloys is lower (8.59.0) than the optimum
pH for carbon and low alloy steels is greater than 9.3 (25C, 77F) (Figures 3-1 and 3-2).

8-7

EPRI Licensed Material


Solutions for Copper Problems

In cycles with copper alloys in feedwater heaters, optimization can only be achieved by
detailed monitoring with different water treatment chemicals at different
concentrations as outlined in the optimization (road map) section later.
Because of the copper alloy corrosion by ammonia and ammonium carbonate, in some
of these cycles, morpholine and/or cyclohexylamine have been used. Other volatile
amines may also be applicable, but their effects on copper alloy corrosion and their
thermal decomposition in fossil cycles need to be evaluated carefully in each case.
Controlling impurity ingress via condenser leaks, and air in-leakage are essential.
Oxygen in synergistic interactions with ammonia and carbonates is very corrosive to
copper alloys. This is particularly pronounced during startups when aerated fill and
feedwater is combined with layup boiler water containing high concentrations of
ammonia and hydrazine. The optimum procedure is to drain the boiler if it contains
large amounts of layup chemicals and refill with normal feedwater before firing.
The most common feedwater treatment is AVT with ammonia and hydrazine. Higher
hydrazine concentrations sometimes reduce the concentration of copper alloy corrosion
products in feedwater. In other very isolated cases, reducing hydrazine feed and even
eliminating all oxygen scavengers during normal operation minimizes the corrosion
products. It is therefore recommended to conduct a detailed monitoring campaign to
determine what is the optimum oxygen scavenger concentration for each specific unit.
AVT with morpholine and/or cyclohexylamineplus an oxygen scavenger is the
second most common feedwater treatment for mixed metallurgy systems. It has been
used in units with copper alloys in feedwater heaters. This water treatment does not
fully optimize feedwater chemistry and prevent steel and copper alloy corrosion, but it
can be used satisfactorily, together with operation and layup steps, particularly in units
with condensate polishers. Morpholine itself works as an oxygen scavenger and,
therefore, use of another scavenger can be minimized, which reduces corrosion of
steels. However, morpholine and cyclohexylamine thermally decompose at normal
fossil plant steam temperatures; the plant cycle must be monitored and these
decomposition products identified.
Use of chelants, polymeric dispersants, and organic oxygen scavengers for optimization
of boiler and feedwater treatment needs to be properly evaluated with respect to their
decomposition, monitoring, analytical interferences, and toxicity. None of these
chemicals are desirable, needed or recommended in utility steam cycles.
Figure 8-1 shows a road map methodology for optimizing feedwater treatment for
mixed metallurgy (copper and iron) feedwater systems where optimum pH control
(8.89.1) is a compromise of the optimum corrosion protection for each metal.
The optimization consists of three general steps:
8-8

EPRI Licensed Material


Solutions for Copper Problems

1. A review of current water treatment, sampling and monitoring, design, operation,


materials, and experience.
2. Implementation of changes when required, including changes of water treatment,
operation, and material replacement.
3. Verification of the effects of the implemented changes. In this verification step. the
water and steam chemistry will be compared to the recommended limits in Table 26 and to other criteria.

Source: Reference 15

Figure 8-1 Road Map for Optimizing Feedwater Treatment for Mixed Metal Systems
(Cu/Fe).
8-9

EPRI Licensed Material


Solutions for Copper Problems

Step 1Review of Water Chemistry, Operation, and Experience


Water Chemistry.
The first part of this review is verification of proper sampling and chemistry
monitoring practices to ensure that valid data are collected; in particular, for
monitoring of iron and copper in feedwater, boiler water and steam, and determination
of drum boiler carryover. The deficiencies must be corrected before Step 2 is initiated.
The chemistry data should be compared to customized EPRI guideline limits shown in
Table 2-6, and a judgment should be made whether they are acceptable or not.

Review of Operation.
The type of operation, layup, and startup procedures should be reviewed for their
relation of cycle chemical transport, particularly their relation to oxidation of copper
alloys and boiler carryover. For cycling and peaking units, the influence of operation
on air in-leakage, layup protection, and transient chemistry should be reviewed.

Experience.
The maintenance reports need to be reviewed and personnel interviewed to evaluate
the results of nondestructive testing and inspections, and the residual life of major
components influenced by water and steam chemistry. Specifically, the status of
condenser and feedwater heater, boiler, and superheater tubing and problems with
flow-accelerated corrosion of carbon steel in the feedwater cycle and in the economizer
will be used in the decision making about design and material changes. It is less
expensive to replace old, corroded feedwater heater tube bundles than good tube
bundles.
If there are no problems such as those indicated in Table 1-2 and 1-3, then continue to
use the current treatment. Such a review would indicate that the operating experience
has been good, that minimal chemical control problems have been experienced, that
minimal boiler tube failures (BTF) in the waterwalls relating to waterside problems and
deposition of feedwater corrosion products have occurred in the last five years, that
there have been no turbine deposition problems or blade failures, and that the
feedwater is operating in the optimum fashion with minimum levels of feedwater
corrosion products (<5 ppb Fe and <2 ppb Cu at the economizer inlet). In such cases of
good experience, no changes need to be made. However it is suggested that the road
map is reviewed as there may be considerable economic savings to be gained.

8-10

EPRI Licensed Material


Solutions for Copper Problems

Step 2Baseline Monitoring


Baseline monitoring is a step which determines cycle chemical characteristics such as
boiler carryover, oxygen control, and transport of corrosion products. During the
baseline monitoring with verified sampling and analytical procedures, compliance with
specifications of key water and steam chemistry parameters is quantified. If problems
exist, this monitoring should also determine where and during what type of operation
there are corrosion and water chemistry problems. For example, which feedwater
heater generates copper alloy corrosion products, tube side vs. shell side, during
startup and/or normal operation, etc.
This program would utilize the installed chemistry monitoring system, supplemented
by monitoring those parameters in the cycle chemistry diagrams under
troubleshooting and commissioning and those in Table 1-1 under Additional
Monitoring or Diagnostic Parameters.
The monitoring program should pay particular attention to the adequacy of the
makeup and chemical feed systems, condenser tightness, air ingress, and corrosion
product transport.
This monitoring usually involves taking a thumb-print of the unit under typical
operating conditions to identify under controlled conditions exactly how the unit
chemistry is behaving. In the case of a serious copper deposition problem on the unit,
this step is key in identifying the root cause of how the copper is entering the steam. It
may involve a review of the operating chemistry logs, but this often is not satisfactory
and it is preferable to adopt a monitoring campaign. Before this campaign is initiated,
it is important to review the utilitys chemistry monitoring capability and reliability (as
in Step 1). This should include Quality Assurance (Q/A) and Quality Control (Q/C) of
existing and normally utilized analytical chemistry monitoring and analysis
methodology and equipment (see EPRIs Cycle Chemistry Monitoring Program(16,17)).
The monitoring campaign should include:

Varying Operating Conditionsbase load, startup, shutdown, etc. It should also


include the effect of drum water level control during startup and transients.

Steam Chemistrycation conductivity, sodium, chloride, silica and sulfate

Feedwater Chemistrycation conductivity, chloride, corrosion products (Fe, Cu),


oxygen, pH, oxidizing-reducing potential (ORP)

Operation of Condensate Polishers (if utilitzed)

During the tests for optimum pH and ammonia concentrations (8.89.1) for mixed
metal system, the pH measurements should be made with on-line equipment designed
for pH measurement in high purity water at carefully controlled sample temperatures
8-11

EPRI Licensed Material


Solutions for Copper Problems

(25C0.1). Ammonia concentrations should be related to specific conductance to be


sure that no carbon dioxide is present.
If this step indicates a low level of feedwater corrosion product transport (such as Fe <5
ppb and Cu <2 ppb) and acceptable feedwater purity from a dissolved solids
standpoint, with the control chemistry meeting the guideline values provided in Table
2-6, then it is suggested to continue with the current chemistry. Further points of
confirmation of optimum feedwater chemistry will be that waterwall deposition rates
are much less than 1 mg/cm2/1000 hours, and that the interval between chemical
cleans has historically been better than every 68 years.

Step 3Water Chemistry Optimization


If water chemistry, and corrosion and scale problems are identified in the review part
of the optimization process, water chemistry optimization can possibly improve the
situation. This step needs to involve operation, startup, and layup. The feedwater
chemistry optimization usually consists of reducing the ingress of impurities, such as
air in-leakage and condenser leakage, minimization of ammonia concentration, and
determination of the optimum oxygen scavenger concentration to minimize copper
corrosion while controlling corrosion and flow-accelerated corrosion of carbon and low
alloy steels. In high pressure drum boiler and once-through cycles, there is little room
for application of alternate oxygen scavengers and volatile amines because of the
problems with their thermal decomposition and generation of organic acids and CO2.
In many units with copper alloys in the feedwater cycle, hydrazine is the only chemical
used because its decomposition provides enough ammonia for pH control. Layup
chemistry and chemical dosing and disposal of the layup solutions must also be
considered in this step.
Thus in this step, a series of tests should be performed to determine the proper
scavenger levels. The tests should utilize the monitoring system instrumentation (Step
2) while reducing the hydrazine (or alternative) dosage. Particular note should be
made of dissolved oxygen levels, quantities of corrosion products, and the
oxidizing/reducing potential (ORP). The latter is particularly important with mixed
metallurgy feedwater systems as a reducing environment must be maintained for the
copper containing alloys as indicated in the Guiding Principles of Section 2.
The choice of oxygen scavenger is also involved here. This includes safety as well as
optimized reaction rates and concern for the alternate behavior of any thermal
decomposition products and their effects in other parts of the cycle. Carbohydrazide,
for example, decomposes to hydrazine and carbon dioxide.
In optimizing copper transport, particular attention must be paid to the fact that there
can be a long time (maybe six months) between changes in feedwater chemistry (ORP)
and corrosion product transport (see Section 7 and Figure 7-3).
8-12

EPRI Licensed Material


Solutions for Copper Problems

Step 4Design and Material Changes


Once design deficiencies or improper materials are identified as the root causes of
water chemistry and corrosion problems, these deficiencies need to be corrected. Such
corrections are usually expensive and a cost-to-benefit analysis will need to be
performed to justify these corrections(1,16). Typical design deficiencies include
feedwater heater inlet design resulting in flow-accelerated corrosion of the tube inlets,
design of drum boiler internals resulting in high mechanical carryover, cycle design
without condensate polishers, and cycle design without pre-boiler cycle cleaning loops.
The material deficiencies include condenser tubing susceptible to steamside or cooling
water side corrosion or erosion, and the use of copper alloys in high pressure feedwater
heaters.

Step 5Operation
Operational changes can be implemented which are not very costly and can reduce
water and steam chemistry problems, such as generation and transport of copper alloy
and carbon steel corrosion products. Other changes can be applied to layup practices
and to system fill and startup. Examples of operational changes include reduction of
boiler pressure during startups when the concentrations of metal oxides are high, more
frequent and longer use of condensate polishers, draining of impure water from
equipment, and blowing down mud drums and lower headers to remove metal oxides.
Each suggested operational change should be tested with the
monitoring/instrumentation as in Steps 2 and 3.

Step 6Monitoring to Compare with Baseline Data


This step is aimed at verifying the effectiveness of the implemented solutions to water
chemistry and corrosion problems. This water chemistry monitoring and
determination of the new cycle chemistry transport characteristics should cover base
load, load variations, startup and shutdown, and it should include all the parameters in
the water and steam chemistry guidelines. Special instrumentation and sampling could
be used to ensure that the corrective measures are effective. There should also be a
long-term monitoring plan which includes equipment inspections and nondestructive
testing.

Step 7Normal Operation and Monitoring


When Step 6 results are satisfactory, indicating long-term, trouble-free water chemistry
and operation, normal operational monitoring resumes. However, there should be
periodic reviews of water chemistry and equipment scale and corrosion problems, and
chemical transport studies should be performed every two to three years to ensure that
8-13

EPRI Licensed Material


Solutions for Copper Problems

equipment aging and other changes do not result in problems. The water chemistry
and operating manuals should be updated to reflect the changes resulting from the
optimization.
The major criterion of a successful treatment for mixed metallurgy feedwater systems is
that the core operating parameters should be within the normal guideline limits (Table
2-6). Also it is important that at the economizer inlet the iron levels should be less than
5 ppb and copper should be less than 2 ppb during continuous operation.

Step 8Continual Check of Chemistry


The continual check on the optimized treatment (Step 8) and the core parameters will
determine whether fine tuning is necessary (Repeat of Steps 3 and 6).

Step 9Long-Term Plans


Based on the effectiveness of the solutions of water chemistry and corrosion problems
implemented in a specific unit and on the equipment life predictions, i.e., feedwater
heater and condenser tube deterioration, long-term plans can be formulated. This can
include planning for ultimate replacement of copper alloys, retrofit of condensate
polishers, nitrogen sparging and blanketing of the storage tanks used to fill the system
during startup, etc.(1)
The industry experience shows that the maximum reduction of corrosion product
transport in fossil utility cycles can only be obtained by eliminating chemical control
compromises caused by the use of mixed metal feedwater systems. Experience has also
shown that the installation of a full-flow condensate polishing system with a startup
clean-up system provides optimized corrosion product transport control.

Solutions 6 and 7. Develop layup/shutdown/startup procedures.


Prevention of corrosion during inactive periods by proper lay-up of heaters and boilers
is an important step in controlling the problems of copper impurities in the feedwater
on unit startup. A lay-up program that maintains an inert environment with ORP < 0
mV) for the boiler and feedwater heaters will reduce copper transport. During layups,
the following should be avoided to minimize copper corrosion:

Exposure of copper alloys to oxidizing conditions (air, aerated water)

Exposure of deposited copper in the boiler and superheater to oxidizing conditions.

Boiling out of the boiler layup water dosed with ammonia and hydrazine.

Filling heater and other components with aerated water.

8-14

EPRI Licensed Material


Solutions for Copper Problems

Ensuring that the oxidizing-reducing potential (ORP) is less than zero millivolts during
all periods of operation and shutdown is one of the Guiding Principles in Section 2.
EPRI is currently preparing shutdown and layup guidelines, which will
comprehensively cover this solution.
C. EQUIPMENT CHANGES

Solution 8. Change Feedwater Heater Materials(20-25)


Replacement of copper alloy tubing in the feedwater heaters is the optimum solution to
the problem and the one which all supercritical and once-through units should have
taken. After detailed cost analysis, some utilities have been able to show considerable
economic benefit associated with retubing of feedwater heaters and condensers(2). In
addition to dealing with the problem of copper and its transport, this has enabled
utilities to switch to oxygenated treatment of feedwater which has provided additional
benefits throughout the cycle.

Feedwater Heater Replacement


A replacement feedwater heater can cost from several hundred thousand dollars to
more than one million dollars. The replacement decision may be: 1) to remove copper
bearing material, 2) because the heater has too many tubes plugged and the forced
outage rate due to tube leakage is not acceptable, 3) due to economic penalties of
copper deposition problems or 4) a combination of all these needs.
There are basically three choices for replacement:
1. Retubing
2. Bundle replacement
3. Complete replacement
The pros and cons of each alternative are discussed in this subsection. It will serve to
help in making the economic choice regarding which alternative is optimum. In
addition, the pros and cons of each replacement non-copper bearing material will be
discussed.

Retubing.
Retubing is the repair process where the old tubes are removed from an existing heater,
and are replaced with new ones. Generally, all other parts of the old heater are reused.
(Internal parts that have become severely damaged due to corrosion or erosion can also
8-15

EPRI Licensed Material


Solutions for Copper Problems

be replaced.) The new tubes must be of the same diameter and length, although
material and thickness can be changed. In the context of this report, it would not be
sensible to replace with a copper alloy.
The advantages of retubing are savings in cost and time. Since all parts, except the
tubes, are being reused the cost is kept to a minimum. Some utilities have performed
their own retubing on site to achieve additional savings. If tubing can be ordered
before taking the heater out of service and retubing can be performed on site,
downtime can be kept to a minimum.
The disadvantages of retubing are:
a. The problems which caused the original heater to fail still exist. Very few
modifications can be done to the heater to correct any existing problems.
b. Tube outside diameter, length, and number of U-bends cannot be changed. If the
tube material is changed the tube thermal conductivity may change and thus cause
a change in heat transfer. If the tube wall thickness has to be changed due to a tube
material change, heat transfer and pressure drop will be affected. A change in heat
transfer and/or pressure drop may cause additional problems within the heater or
in the equipment connected to it.
c. High pressure feedwater heaters usually require welded tube to tube sheet joints for
greater reliability. To achieve a good tube to tube sheet joint, tube sheet weld
preparation and cleanliness are very important. As a rule, it is considered
impractical to reweld tube to tube sheet joints after retubing. If the tube material
has been changed the overlay material on the tube sheet will most likely need to be
changed in order to properly weld the tube joint. Generally, it is considered an
insurmountable task to overlay a tube sheet which has holes already drilled in it.
d. If a tube hole has grooves or worm holes in it, it may not be practical to install a new
tube and still produce an acceptable rolled tube joint. In this case the hole will have
to be plugged. If there are too many holes in this condition the tube sheet may have
to be replaced, causing a considerable delay in returning the heater to service.

Bundle Replacement.
Rebundling is the rebuilding of a heater using new tubes, a new tube sheet and shell
skirt, and reusing the part of the shell beyond the cut line. Other internal parts may or
may not be reused depending on their condition and whether or not there is a change
in design. The channel of a low pressure heater may be reused. A hemispherical head
from a high pressure heater may also be reused, but generally the channel of a high
pressure heater with a full access closure cannot because the tubesheet is an integral
part of the channel.
Rebundling is probably the most economical way to repair and correct most of the
problems of a malfunctioning heater. The cost savings may be from 5 to 20% of the cost
of a complete replacement heater. The amount of savings depends on the size, design
8-16

EPRI Licensed Material


Solutions for Copper Problems

pressures on the shell side and tube side, replacement tubing material, etc. The
advantages of repairing a heater this way are:
a. The replacement bundle may be ordered while the heater remains in service. When
the replacement bundle is received, the old bundle may be pulled out of the shell
and replaced with the new one. The existing shell may be left in place while doing
this and none of the shell connections need to be disturbed.
b. Since the tube sheet will be replaced, any problems encountered in retubing the old
tube sheet will be eliminated. Tubes can be welded to the tube sheet even if the
material has been changed.
c. Tube size, thickness, material, layout pattern and internal design can be changed to
correct any problems the original heater encountered. Frequently, additional heat
transfer surface can be provided without increasing the length of the heater by
utilizing smaller outside diameter tubes.
In order to reuse the original shell, it must, of course, be in reusable condition. It is of
utmost importance to know the exact inside diameter of the shell to avoid any problem
in inserting the new bundle. Many utilities prefer to ship the old shell to the bundle
manufacturer where the new bundle is then inserted. The new bundle is also protected
during shipment in this way. It is also necessary to know the exact inside diameter of
the original hemispherical head if it is to be reused.

Complete Replacement.
In complete replacement a new heater is designed and built using all new parts.
Obviously a complete replacement will cost more than retubing or rebundling. A
complete replacement offers all the advantages of a rebundle plus:
a. A new replacement heater may be ordered and placed into service at the most
opportune time. Since no part of the old heater is reused, the old heater may be left
in service until the last minute.
b. It allows a change in shell diameter which may be necessary to rectify the problem
of the original heater.
c. Potential fit-up problems are eliminated if the inside diameter of the original shell
cannot be accurately determined.
In rebundled and complete replacement units, all major original piping terminal point
locations can, with few exceptions, be maintained. Minor connections, such as venting
and liquid level controls, may have to be relocated due to internal design changes.
When the type of channel closure is changed, for instance from hemispherical to full
access, the tube side connections will most likely have to be changed.
It is always helpful to provide the feedwater heater manufacturer with thermal,
operating, and mechanical design data as well as a sketch showing accurate locations of
8-17

EPRI Licensed Material


Solutions for Copper Problems

all major connections. Much information, such as nominal inside diameters, materials,
and thicknesses may be listed on the U-1 form (Manufacturers Data Report for Pressure
Vessels). These may be found in the original heaters instruction manual and/or
quality assurance data report. U-1 Forms may be obtained by writing to the National
Board of Boiler and Pressure Vessel Inspectors, 1055 Crupper Avenue, Columbus, Ohio
43229-1183 Tel. (614) 888-8320. The National Board registration number can be found
on the feedwater heater nameplate.
In evaluating the possible cause of failure of the original heater, it is helpful to know
the history of the heater and to have a map of plugged tubes. In order to redesign a
bundle or heater to avoid possible problems, it is helpful to provide the manufacturer
with operating information for loads other than the design point, such as partial or
overload conditions, and to also specify whether upstream heaters can be bypassed and
whether the heater will be subjected to cyclic service.
Whichever replacement method is used, retubing, rebundling or complete replacement,
the feedwater heater manufacturer should be able to provide the best service. Take the
maximum advantage of their knowledge and experience, by providing them with the
best available information, including the problems in the existing design which should
be addressed and corrected.

Feedwater Heater Tube Material Comparison


The most critical part of any closed feedwater heater(FWH) is unquestionably the tube.
This subsection will examine some of the basic characteristics of tubes and attempt to
briefly outline the effects of employing various kinds of tubing material in FWHs. The
list of tube materials includes copper bearing material to show the effect of changing to
non-copper alloy tubing.
The difference between tubes in the H P and the LP, aside from the fact that they
operate at high and low pressure, is that, as a general rule, the tube wall thickness of
tubes used in the LP is set NOT by the operating conditions, but, rather, by either the
user or Heat Exchange Institute (HEl) Standards (Table 8.4). This is because the usual
LP tube side design pressures and temperatures are low enough (less than 600700 psi
and under 400F) that the wall thicknesses required by HEI are higher than those
necessary to meet the American Society of Mechanical Engineers (ASME) Code.

Table 8-4
Design Minimum Tube Wall Thickness
Copper and Copper Alloy

.049 avg. wall

Nickel Alloy

.049 avg. wall

8-18

EPRI Licensed Material


Solutions for Copper Problems

Table 8-4
Design Minimum Tube Wall Thickness
Stainless Steel (U-Tubes)

.035 avg. wall

Stainless Steel (Straight Tubes)

.028 avg. wall

Carbon Steel

.050 avg. wall

In view of this, the following discussion will be directed at HP tube materials noting
that, ignoring strength considerations, whatever is said about HP tubing would
generally also apply to LP tubing.
When tube material and ASME tube specification are decided, the net effect is that the
heat transfer capability and the strength capability of the tube is set. From then on, the
design parameters and the various governing documents control.
Table 8-5 lists the more commonly used tube materials, the ASME Code allowable
stress, the thermal conductivity (this is the measure of ability of the tube material to
transfer heat), and ratios which can be used to compare the various materials. It should
be noted that cost is not considered in this section, but will be discussed in the next
section.
Note that the materials are arranged in relation to strength (strongest at the top), and
that the relative stress, S/Sm, and relative conductivity, k/km, are the ratios of the
stress and conductivity of the material to that of Monel.

8-19

EPRI Licensed Material


Solutions for Copper Problems

Table 8-5
Feedwater Heater Tubing and
Properties
S@

S/Sm

k/km

HEI

Material

400F

s Monel

21.0

1.0

16

1.0

10

s AL29-4C
s Seacure

20.9

.995

9.8

.613

10

w 304N SS

18.3

.871

10

.625

10

w AL29-4C
w Seacure

17.8

.848

9.8

.613

10

s CS-Grd C2

17.5

.833

28

1.75

s 70-30 Cu-Ni

17.3

.824

21

1.313

s 304SS

16.2

.771

10

.625

10

w 304LN SS

16.2

.771

10

.625

10

15

.714

21.5

1.34

s 304L SS

14.7

.7

10

.625

10

w 304 SS

13.8

.657

10

.625

10

s 439 SS

13.3

.633

15

.938

10

w CS-Grd B

12.8

.610

28

1.75

w 304L SS

12.5

.595

10

.625

10

w 439 SS

11.3

.538

15

.938

10

s 90-10 Cu-Ni

8.5

.405

34

2.125

w Titanium

10@300F

.476

12.5

.781

10

s Admiralty

10@300F

.476

79

4.938

8.5

s Cu (Lgt. Drwn.)

8.7@300F

.414

224

14

8.5

w T-22

VHEI

Where:
CS = Carbon Steel
k = Thermal Conductivity Btu/hr ft F (1 Btu/hr ft F = 1.7307 W/m C)
m = (subscript) = Monel
S = ASME Code allowable stress, ksi
s
= Seamless
SS = Stainless Steel
w = Welded
V = HEI Maximum Velocity, fps (1 fps = 0.305 mps)

For tubes of different materials, but of the same thickness, the ratio k/km provides a
fast way to assess the ability of the tube to transfer heat.
8-20

EPRI Licensed Material


Solutions for Copper Problems

For HP tubing the relative thermal conductivity does not totally describe the heat
transfer of a material since, for the same pressure and temperature, the wall thickness
required to resist the pressure increases as the strength decreases. This wall thickness
increase causes a reduction in heat transfer.
Another factor in choosing feedwater tubing is the allowable velocity of the feedwater
in the tubes. HEI has established velocity limits for the various materials and these are
listed in the last column of Table 8-5. Some of the effects on a heater as the allowable
velocity increases are increased pumping power requirements (provided tube lengths
are about the same), higher tube side heat transfer coefficients, and a reduced number
of tubes required per pass. It should be noted that these are general trends which take
place; however, the particular set of conditions which have to be met for each job may
be such that other factors, e.g., maximum heater length, bundle withdrawal space, etc.,
govern the final design.
It can be seen that the maximum velocity for 90-10 Cu Ni tubes is 9 fps, whereas for
Monel tubes the maximum velocity is 10 fps. This requirement, coupled with the
heavier wall thickness required for 90-10 Cu Ni tubes, because of its lower strength,
would necessitate using about 1.9 times as many 90-10 Cu Ni tubes as Monel tubes in
order to satisfy the design criteria for any particular heater. Depending on the
circumstances, the increased number of heavier 90-10 Cu Ni tubes could:
1. make the cost of the heater with 90-10 CuNi tubes exceed the cost of a heater with
Monel tubes even though the Monel is more expensive on a $/# basis, or
2. make the remainder of the heater design impractical, e.g., cause the diameter to get
so large that the metal thicknesses required to resist the pressure become absurd.
For low pressure feedwater heaters, the first choice for replacement tubing material is
welded 304 stainless steel, which has good corrosion resistance and it is relatively
inexpensive. The tubing should be ordered with 0.05% maximum carbon and the
residual stress of the tubes, after straightening, or straightening and bending, should
not exceed 500 psi circumferentially (hoop stress) and 300 psi axially, in either tension
or compression. When retubing a 90-10 Cu-Ni feedwater heater with 304 stainless, the
difference in performance degradation will be undetectable. If the original tube
material is admiralty, replacing with welded 439 s.s. will reduce the performance lost.
304 s.s. and 439 s.s. are susceptible to stress corrosion cracking and pitting in the
presence of chloride. If cost is not a factor, the best replacement tube material is
titanium. Although not many heaters are built with titanium, it offers the best in heat
transfer and corrosion resistance. For a complete evaluation of replacement tube
material, the user should use the EPRI Heat Exchanger Workstation - Feedwater Heater
Application software(23).
For high pressure feedwater heaters, the most popular choice for replacement tubing
material is welded or seamless 304N stainless steel. AL29-4C, Seacure, and 439 s.s.
should not be used if the steam temperature can exceed 600F because the possibility of
8-21

EPRI Licensed Material


Solutions for Copper Problems

885 embrittlement. To avoid the possibility of stress corrosion cracking, super stainless
steel such as AL6XN may be a better choice. Carbon steel could also be used, however,
its useful life may be limited to about 7 years due to corrosion and erosion. Recently,
several utilities have used low alloy steel such as T-22, which is relatively inexpensive
and offers good strength and considerably better corrosion and FAC resistance than
carbon steel. Again for a complete evaluation of replacement tube material, the user
should use the EPRI Heat Exchanger Workstation - Feedwater Heater Application
software(23).

Examples of Replacing Tube Material in Feedwater Heaters


This subsection will discuss, investigate, and provide some examples on the effect of
using different tube materials in feedwater heaters.
The comparisons discussed are based on a coal fired 800 MW unit with a dual
feedwater heater train. The high pressure heater is a three zone heater designed for
3,500 psig at 400F, having a hemispherical head closure with welded tube-to-tube
sheet joints. The tube O.D. is .625 in, with the velocity inside the tube being 8 fps. The
low pressure heater is a two zone heater designed for 300 psig at 350 F, having a bolted
cover plate closure with rolled, but not welded, tube joints. The tube O.D. for the low
pressure heater is .75 in, and the velocity inside the tubes is 6 fps. Other design
characteristics are in accordance with the HEI Standard. The heaters with carbon steel
tubes include stainless steel inserts and diffuser plates at the tube inlet end. Welded
TP-304 stainless steel tube is used as a base for comparison.
In all cases, the temperatures, flows, and velocities are constant. The configuration of
the heater is changed to maintain constant tube side velocity.
The following factors (Tables 8-6 and 8-7) need to be considered when replacing
feedwater heaters:

8-22

EPRI Licensed Material


Solutions for Copper Problems

Table 8-6
LP Feedwater Heater Comparison Chart

Welded

Tube Wall

Tube Side

Metal

Surface

Thk.

Flow Area

Resist.

Factor

Shell Dia. No. of Tube

Length

Pressure

Cost of

Cost of

Factor

Drop Factor

Tubing

Heater

Factor

Factor

Factor

Factor

0.35A

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

0.35A

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.06

1.03

0.35A

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.45

1.19

0.35A

1.0

.667

.941

1.0

1.0

.941

.941

1.45

1.19

0.49A

.919

.062

.848

1.043

1.09

.779

.868

1.28

1.10

.049A

.919

.190

.872

1.043

1.09

.801

.871

1.24

1.09

.049A

.919

.461

.920

1.043

1.09

.846

.92

1.62

1.25

Factor

Factor

304
S.S.
Welded
304L
S.S.
Welded
316
S.S.
Welded
439
S.S.
Smls.
Copper
Smls.
Adm.
Smls.
90-10
Cu Ni
A = Average Wall
Numbers are based on tube O.D. of .75 and a velocity of 6 ft./sec.

8-23

EPRI Licensed Material


Solutions for Copper Problems

Table 8-7
HP Feedwater Heater Comparison Chart
Tube Wall

Tube Side

Metal

Surface

Thk.

Flow Area

Resist.

Factor

Factor

Factor

1.25

.406

Smls.

0.58A

Monel

.058M

.681

Shell Dia. No. of Tube


Factor

.894

Factor

.799

Length

Pressure

Cost of

Cost of

Factor

Drop Factor

Tubing

Heater

Factor

Factor

3.76 (D)*

1.99

2.18 (F)*

1.36

.853

.762

DSR
Smls.

.065M

1.13

.291

.642

.942

.887

.724

.682

.554

.772

0.65M

1.15

.375

.680

.934

.873

.779

.728

1.01

.898

Smls.

0.72A

1.11

.844

.91

.951

.904

1.01

.958

1.82

1.29

304 S.S.

0.72M

Welded

0.72M

1.06

.915

.95

.973

.947

1.0

.976

1.53

1.19

Welded

0.83A

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

304 S.S.

0.83M
.943

1.09

1.05

1.03

1.06

.993

1.02

1.22

1.12

.840

.85

.958

1.09

1.19

.804

.877

1.68

1.43

Carbon
Stl
Gr C2
Smls.
70-30
DSR

316 S.S.

Welded
304L

0.83M

S.S.

0.95A

Welded

0.95M

439
S.S.
*D = Domestic
A = Average Wall
F = Foreign
M = Minimum Wall
Numbers are based on tube O.D. of .625 and a velocity of 8 ft./sec.

8-24

EPRI Licensed Material


Solutions for Copper Problems

Tube Wall Thickness


The tube wall thickness is for high pressure feedwater heaters and is based on the HEI
formula and the ASME Code allowable stress at 400 F. ASTM manufacturing tolerances
on all wall thicknesses, and normally produced thicknesses, have been considered in
arriving at the selected wall thickness. When there are two different wall thicknesses
listed, the inner bends use the heavier wall tube. In Table 8-5 for low pressure
feedwater heaters, the allowable stress of the tube material is not a factor in arriving at
the tube wall thickness in the design. HEI recommends a wall thickness of 20 BWG
average wall for stainless steel, and 18 BWG average wall for copper alloy.
Tube Side Flow Area
This factor is based on the tube I.D. after the tube wall manufacturing tolerances have
been taken into account It should be noted that in the case of minimum wall tubes, this
results in a more conservative design than that which would be arrived at if nominal
dimensions were used. All manufacturers do not use this approach. This factor
influences the heater diameter and number of U tubes when based on a constant tube
side velocity.
Metal Resistance
This factor is based on the thermal conductivity and wall thickness of the tube material
selected. Again, the wall thickness tolerance is taken into consideration.
Surface
This factor is based on the surface required to transfer the same amount of heat at the
same temperature differential. Metal resistance, tube side velocity, and HEI specified
minimum fouling factors are also considered. In order to simplify calculations, only
the condensing zone of the heater is considered since 75% to 80% of the heat transfer
surface is in the condensing zone, and the desuperheating zone and subcooling zone
heat transfer rates do not change appreciably with change in tube material and wall
thickness. The maximum error introduced by these assumptions is 2%.
Shell Diameter and Number of Tubes
The outside diameter of the heater and the number of U tubes in a heater are a function
of the tube wall thickness when the velocity through the tubes is held constant
Length
This is a function of the total surface required and the number and O.D. of the tubes in
the heater.

8-25

EPRI Licensed Material


Solutions for Copper Problems

Pressure Drop
Pressure drop on the tube side is a function of the tube inside diameter, length of
heater, and velocity. This could be an important factor when considering feedwater
pumping power consumption.
Cost of Tubing
This factor is based on the cost of tubing and forming of U bends.
Cost of Heater
This is a function of tube cost, other material costs, and labor. In this factor, diameter,
length, and number of tubes need to be taken into consideration.
The following two examples provide information on changing feedwater heater tubing
from a copper base alloy to stainless steel.

Example A, Low Pressure Feedwater Heater


The original low pressure feedwater heater had admiralty brass tubes which will be
replaced with 304 stainless steel tubes.

Table 8-8
Key Factors Involved in Replacing Admiralty Tubes with Stainless Steel in a LP
Heater (Data extracted from Tables 8-6 and 8-7)
Item

Factor

Ratio

Metal Resistance

1.0/.19

5.26

Surface Area

1.0/.872

1.15

Shell Diameter

1.0/1.043

.96

Number of Tubes

1.0/1.09

.92

Length

1.0/.801

1.25

Pressure Drop

1.0/.871

1.15

Cost of Tubing

1.0/1.24

.81

Cost of Heater

1.0/1.09

.92

8-26

EPRI Licensed Material


Solutions for Copper Problems

From Table 8-8, it is evident that 304 S.S. has a higher metal resistance than admiralty
and, therefore, under the same velocity and temperature conditions, more surface will
be required for the same duty. This means, then, that the low pressure feedwater
heater with 304 S.S.tubes, in order to handle the same duty with the same flows, will
require 15% more surface, be 25% longer, and have 15% more pressure drop. 304 S.S.
tubes will cost approximately 19% less than admiralty, the necessary increase in tubing
length means that the heater will only cost approximately 8% less.
If the greater tube length and pressure drop cannot be tolerated, it may be necessary to
increase the number of tubes to reduce the pressure drop. However, surface area and
other factors cannot be arrived at by direct comparison and, therefore, the heater
manufacturer should be contacted to conduct a detailed study.

Example B - High Pressure Feedwater Heater


The original high pressure feedwater heater had 70-30 Cu Ni tubes, which will be
replaced with 304 S. S. tubes. From Table 8-5 stainless steel has a lower allowable stress
and lower thermal conductivity than 70-30 Cu Ni. Therefore, heavier wall tubing and
more surface will be required for the same duty.

Table 8-9
Key Factors Involved in Replacing 70-30 Cu-Ni Tubes with Stainless Steel in a
HP Heater. (Data extracted from Tables 8-6 and 8-7)
Item

Factor

Ratio

Metal Resistance

1.0/.375

2.67

Surface Area

1.0/.680

1.47

Shell Diameter

1.0/.934

1.07

Number of Tubes

1.0/.873

1.15

Length

1.0/.779

1.28

Pressure Drop

1.0/.728

1.37

Cost of Tubing

1.0/1.01

1.0

Cost of Heater

1.0/.898

1.11

This comparison (Table 8-9) indicates that approximately 15% more tubes are required
for the same tube side velocity and, therefore, the shell diameter can be increased
8-27

EPRI Licensed Material


Solutions for Copper Problems

approximately 7%, with the surface area increased approximately 47%. It should be
noted that the replacement heater will be approximately 28% longer and, as a result,
the pressure drop will be approximately 37% higher. Since the S.S. tubes cost
approximately the same as Cu Ni tubes, a heater with S.S. tubes will be increased only
11% over the cost of a heater with Cu Ni tubes.
If the greater length and pressure drop cannot be tolerated then the number of tubes
may have to be increased to make the heater shorter.

Solutions 9-11. Modify Unit and Operating Procedures


Use of condensate polishers and condensate/feedwater filtering systems, particularly
during startups can be critical. Of particular importance is the start-up period; the peak
in copper transport usually occurs approximately 1-2 hours into startup (see Section 5
and Figures 5-15-3). Maximum effort should be given to identifying the mode of
operation responsible, and modifying the operating procedures to minimize transport
of copper accumulated during the shutdown period. Such procedures might typically
be:

Introduction of non-aerated makeup water during startup.

Use of startup filters such as using side stream condensate filtration.

Blowdown of the boiler to avoid high copper levels at pressures below 15.817.2
MPa (23002500 psi) during the startup sequence, continue to blowdown until
copper levels have been reduced to < 50 ppb or some reasonable level determined
for the particular characteristics of each unit.

Routing feedwater after the HP feedwater heaters to condenser on startup until


drain copper levels are below 40 ppb.

Nitrogen blanketting the boiler and feedwater heaters during shutdown.

Plants having condensate polishers can significantly reduce corrosion-product


ingress through the installation and use of condensate and feedwater cleanup
loops(6,18,19). Such loops can (i) clean up the hotwell by recirculating condensate
polisher effluent back to the hotwell, (ii) clean up the low-pressure feedwater
heaters by recirculating the feedwater from the deaerating heater back to the
hotwell, and (iii) clean up the high-pressure feedwater heaters by recirculating
feedwater from the economizer inlet back to the hotwell.

It has also been suggested that a copper loading curve can be considered. (See
Section 6 and Figure 6-3.) This is a protocol by which the startup process is delayed
until certain specified levels of Cu are attained via one of the methods outlined
above.

If the copper deposition problem is being caused by excessive mechanical carryover as


a result of poor drum level control or damage to the drum furniture or steam
8-28

EPRI Licensed Material


Solutions for Copper Problems

separation equipment, appropriate actions should be taken. In the former case,


operating procedures should be instituted to eliminate drum level control problems; in
the latter, testing to determine whether there are defects present and repair should be
performed.
REFERENCES
1. Cycle Chemistry Improvement Program . Electric Power Research Institute, Palo Alto,
Calif. TR-106371. April 1997.
2. D. Gunn, D. McInnes, K. Knights and D. Ryan. Combined Oxygen Hydrazine
Ammonia Conditioning (COHAC) and Copper Transport Through the Cycle.
Interaction of Non-Iron Based Materials with Water and Steam. Edited by B. Dooley and
A. Bursik. EPRI Proceedings. TR-108236. July 1997.
3. T. Gilchrist. Hydrazine Feed and its Impact on Corrosion Product Transport at
Craig Station. Fifth Annual Southwest Chemistry Workshop , St. Johns, Arizona. July
1996.
4. Guidelines for Chemical Cleaning of Fossil-Fueled Steam Generating Equipment. Electric
Power Research Institute, Palo Alto, Calif. TR-102401. June 1993.
5. P. Cohen, Editor. The American Society of Mechanical Engineers. The ASME
Handbook On Water Technology For Thermal Plant Systems. New York City, NY, 1989.
EPRI GS-6303, pp. 134, 173, 1336.
6. B. T. Hagewood. Copper Fouling Northport 4 High-Pressure Steam Turbine. 2nd
Conference On Cycle Chemistry In Fossil Plants. Electric Power Research Institute.
GS-6166. January 1989.
7. K. N. Thomsen and K. Daucik. Deposits on Turbines in Relation to the Use of
Copper Alloy Materials in the Water/Steam Cycle. Interaction of Non-Iron Based
Materials with Water and Steam. Edited by B. Dooley and A. Bursik. EPRI
Proceedings TR-108236. July 1997.
8. B. Axley, D. W. Beavers, S. O. Hilton, and M. G. Sexton, Turbine Copper Deposits
in a Supercritical Once-through Unit, in R. B. Dooley, ed., Proceedings: International
Conference on Fossil Plant Cycle Chemistry . TR-100195, Electric Power Research
Institute, Palo Alto, Calif. 1991, pp. 265 through 280.
9. S. J. Cioffi, In-Place Chemical Cleaning of Will County Unit 4 High Pressure
Turbine for Efficiency Recovery, in D. M. Rasmussen, ed., Latest Advances in Steam
Turbine Design, Blading, Repairs, Condition Assessment and Condenser Interaction, 1989
Joint Power Generation Conference, PWR Vol. 7, American Society of Mechanical
Engineers, New York, 1989.
10. B. T. Hagewood, Copper Fouling Northport 4 High-Pressure Steam Turbine, 2nd
Conference on Cycle Chemistry in Fossil Plants, Seattle, Washington, Electric Power
Research Institute. GS-6166. January 1989.

8-29

EPRI Licensed Material


Solutions for Copper Problems

11. P. LaLena and G. Glaser, Consider Foam Cleaning for Turbines, Power, March,
1980, pp. 6668.
12. E. S. Fridley, C. D. Foster, and D. J. Shomo, Chemical Cleaning Steam Turbine
Restores Output, Slashes Cost, Power, July 1990, pp. 5153.
13. Iatan High Pressure Turbine Foam Cleaning, presented at the Spring Meeting of
the Missouri Valley Association, 1990.
14. L. Bursik, W. Clary and A. Bursik. Chemical Cleaning of Components Focusing on
Copper and Copper Containing Deposits. Interaction of Non-Iron Based Materials
with Water and Steam. Edited by B. Dooley and A. Bursik. EPRI Proceedings. TR108263. July 1997.
15. Selection and Optimization of Boiler Water and Feedwater Treatments for Fossil Plants.
EPRI TR-105040. March 1997.
16. Cycle Chemistry Corrosion and Deposition: Correction, Prevention, and Control. Electric
Power Research Institute, Palo Alto, Calif. EPRI TR-103038. December 1993.
17. Monitoring Cycle Water Chemistry in Fossil PlantsVolumes 13. Electric Power
Research Institute, Palo Alto, Calif. EPRI Report GS-7556. 1991/2.
18. F. Gabrielli, N. C. Mohn, and W. R. Sylvester, Water Chemistry Aspects of Cyclic
Operation for Older, High-Pressure Drum-Type Boilers, Proceedings of the American
Power Conference, Volume 45, 1983.
19. D. M. Sopocy and Y. H. Lee, Corrosion Product Control Measures for Existing
Stations, Proceedings of the American Power Conference, April 1985.
20. Feedwater Heaters: Replacement Specification Guidelines , EPRI GS-6913, August 1990.
21. Feedwater Heaters Maintenance and Repair Technology: Reducing Outage Cost , EPRI GS6935, August 1990.
22. Enhanced Reliability of Replacement Feedwater Heaters, EPRI TR-103940, June 1994.
23. Heat Exchanger Workstation - Feedwater Heater Application Users Manual , EPRI SW104591, January 1995.
24. Standards for Closed Feedwater Heaters, 5th Edition, Heat Exchange Institute.
25. ASME Boiler and Pressure Vessel Codes , Section II, Section VIII, American Society of
Mechanical Engineers.

8-30

EPRI Licensed Material

9
ASSESSMENT AND CORRECTION OF COPPER
CORROSION, TRANSPORT AND DEPOSITION
PROBLEMS

INTRODUCTION
The previous sections have provided the state of knowledge about each aspect of
copper, copper alloys and copper oxides in the fossil plant cycle: corrosion,
electrochemistry and transport in feedwater; volatility and solubility in steam; and
deposition on HP turbine components. Section 8 has delineated the main solution
options with particular emphasis on chemical cleaning of turbine components,
optimization of feedwater chemistry, and changing copper-based feedwater heaters to
all-ferrous materials. A set of Guiding Principles for successful operation of units with
copper alloys without major turbine problems has been presented in Section 2. The
EPRI guideline limits for feedwater in mixed-metallurgy systems is provided in Table
2-6.
Clearly as outlined in each of the previous sections, there are deficiencies in the current
knowledge which need to be addressed by future R&D. However, the purpose of this
current section is to bring the information, inherent in the previous sections together, to
develop a logical road-map approach which utilities can use to determine if, or
whether, significant copper problems exist, or might occur in the future. The approach
can, and should, be used by utilities without current problems to improve the specific
conditions in units with mixed metallurgy feedwater systems.
OVERVIEW OF UTILITY ACTIONS
Personnel responsible for performance and chemistry of fossil plant units should
implement several actions in order to effectively assess and correct known or suspected
problems involving copper transport. Figure 9-1 is a road map which identifies
questions that need to be addressed and indicates a generic, sequential structure within
which assessment and corrective measures may be implanted. Depending on the
situation which exists in a specific unit, it may be possible to either implement some
actions in a different sequence than indicated or implement actions in parallel.
9-1

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

The activities indicated in Figure 9-1 should be regarded as a set of interrelated actions.
In general, individual actions begin with an evaluation of existing conditions, which
should be followed by an assessment and (where feasible) implementation of
improvements if the status quo is found to be deficient. Implementation of justifiable
improvements should always be followed by a monitoring campaign, an action
intended to determine how any changes enacted influence copper transport activity
and associated problems in the unit being evaluated.
Actions which need to be considered in copper transport assessments include:

Action 1 - Determine Unit Susceptibility to Copper Problems

Action 2 - Evaluate Condition of Copper Alloy Components

Action 3 - Characterize Copper in Cycle Deposits

Action 4 - Optimize Cycle Chemistry - Monitoring Campaign

Action 5 - Improve Startup Procedures and Layup Practices

Action 6 - Examine Cycle Design

Action 7 - Follow-up Monitoring Campaigns

Guidance and details regarding each of the seven actions are provided in the report
subsections which follow.

Action 1 - Determine Unit Susceptibility to Copper Problems


Action 1 consists of two parts. The first step to be undertaken is Action 1A, to
determine if the unit has experienced any of the problems commonly associated with
excessive copper transport activity. The second step, Action 1B is to determine if the
unit has design characteristics and features which increase the likelihood that copperrelated problems will eventually develop. Together with the major problems in Table
1-3, the following situations should be regarded as examples of problems which may
arise as a consequence of copper transport activity. They can also be regarded as
indicators that the feedwater chemistry is not optimized.

9-2

Turbine capacity losses caused by copper deposits.


Turbine stage efficiency losses caused by copper deposits.
Boiler tube failures involving waterside deposits rich in copper(1).
Waterwall chemical cleanings typically require inclusion of a separate copper
removal solvent.
Boiler feed pump performance losses due to copper deposits in feedwater heaters
and piping.
Excessive deposition of copper on strainers, screens, valves, orifices, etc., which
require maintenance activity.

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Action 1
Do Significant Copper
Transport Problems Exists?

Is Probability Of Future
Problems High?

Action 2
Are Copper Components
Considered Reliable?
Perform Conditions Assessments,
Replace With Copper-Free
Materials When Justified.

Include Copper In Future


Monitoring Campaigns If Present In Heaters Or Condenser.

Action 3
Are Deposits Rich In
Copper Present?

Action 4

Investigate And Evaluate;


Clean When Justified.

Has Unit Cycle Chemistry


Been Optimized?
Optimize And Customize
Feedwater And Boiler Water
Treatment Program.

Action 5
Can Startup Practices
Be Improved?
Evaluate And
Modify Procedures.
Action 5
Are Equipment Layup
Procedures Adequate?

Establish Or Improve
Procedures, And Apply Them.

Action 6
Could Cycle Design
Changes Help?

Action 7

Evaluate Options,
Implement Where Justified.

Does Monitoring Indicate


Effective Control Of
Copper Transport?
Reevaluate Periodically
And If Transport
Problems Return.

Figure 9-1 Summary Road Map for Utility Assessment and Correction of Problems
Involving Copper Transport
9-3

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

In some instances, there may be performance losses in which copper or other


deposition could be involved but the root cause has not been verified. Further work
(inspection of affected component, collection and analysis of deposit samples, etc.) may
be required to determine if the losses are due to copper transport effects. Performance
of steam path audits should be carefully considered whenever turbine performance
appears to be affected by copper deposits.
Any of the following characteristics which apply to the unit in question increases the
probability of future problems related to excessive copper transport. It is assumed that
copper alloys are or were originally in use in low pressure and/or high pressure
feedwater heaters. However, the criteria also apply to units with ferrous heaters and
copper alloy condensers.

Units with drum pressures of 2400 psig and above may experience turbine fouling
as a result of volatile carryover.

Units which are put into overpressure operation

Units with high mechanical carryover rates (in excess of 0.3 percent) may
experience turbine fouling at times when boiler water copper levels are elevated.

Units with high attemperation spray rates (above 5 percent of steam flow) are prone
to turbine fouling at times when feedwater copper levels are elevated.

Units subject to peaking and/or cycling service are prone to increased copper
transport activity relative to baseloaded units; effective layup procedures, which
maintain a reducing environment (ORP < 0 mV) can reduce copper transport.

Units with deficient control of air inleakage and dissolved oxygen are subject to
increased release rates for copper. Units which dont have reducing feedwater
conditions (ORP < 0 mV) during all periods of operation and shutdown will have
increased transport of cupric oxide (CuO).

Units which dont apply nitrogen blankets to the boiler and feedwater heaters
during shutdown.

Units which have not optimized cycle chemistry are more likely to experience
significant copper transport activity. Specifically units that have copper levels
above 2 ppb at the economizer inlet during normal operation will be affected.

Units where the oxygen scavenger has been reduced, eliminated or changed.

Unit designs which do not include condensate filters and/or polishers, and
feedwater cleanup loops are more likely to develop problems involving deposition
of copper.

Unit designs which do not include turbine bypass lines are more likely to
experience fouling problems.

9-4

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Units which include boilers of high heat flux design are more likely to experience
problems involving copper in waterwall deposits.

These activities are schematically presented as road maps in Figures 9-2 (Action 1A)
and 9-3 (Action 1B). In Figure 9-2, it is assumed that copper has been verified as
present in system deposits. Where this is not confirmed, utility personnel should take
appropriate steps to do so. Documentation of costs should include any and all
expenses directly attributable to problems encountered as a result of copper transport
activity. Where appropriate, the costs associated with cleaning or other activities
initiated to restore performance losses should be included.
Collection of cost information is essential in assessing the extent to which
implementation of corrective actions is justifiable. Conclusions of economic evaluations
will naturally be unique for each individual fossil plant unit appraised.
The Action 1B road map (Figure 9-3) serves as a checklist of unit characteristics and
deficiencies which are viewed as likely to lead to problems related to copper transport.
These are taken directly from the Guiding Principles in Section 2. Units which appear
susceptible to copper transport problems should be evaluated further. Units which
meet any of these criteria but are not subject to copper problems should be reevaluated
as necessary to consider the effect of changes in design, operation and chemistry on
transport activity.

9-5

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Figure 9-2 Road Map for Determination of Unit Susceptibility to Copper Problems:
Action 1A, Review of Experience

9-6

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Figure 9-3 Road Map for Determination of Unit Susceptibility to Copper Problems Action 1B, Review of Design and Operating Characteristics

9-7

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Action 2 - Evaluate Condition of Copper Alloy Components


Action 2 , summarized in Figure 9-4, provides direction for assessment of the condition
and future serviceability of copper alloy components in the steam-water cycle. Its use
is intended primarily for feedwater heaters but the concepts may be applied to
condensers or other copper components thought to represent significant sources of
copper release.
The process by which Action 2 is conducted first involves gathering and reviewing
available information on tube leaks and plugging, nondestructive evaluations, tube
sampling and metallurgical evaluation activities, mechanical and chemical cleanings
and, where applicable, component modifications such as tube replacement. If the
historical review determines that failures have occurred but without the necessary
follow-up to establish the damage mechanism and root cause, nondestructive testing,
supplemented by metallurgical evaluation should be implemented to determine this
and estimate the remaining reliable service life. Acoustic testing and eddy current
testing are the preferred methods for nondestructive evaluation of heater tubing.
Action 2 activities require the continued attention of utility personnel as long as copper
alloys are in use in the component of interest. Periodic evaluation is necessary despite
initial indications that no failures have been experienced or that the remaining reliable
service life is satisfactory. Follow-up action is also necessary in cases where the
original copper alloy is replaced with any material which includes copper as a
significant constituent.

Action 3 - Characterize Copper in Cycle Deposits


Based on the information developed in Sections 3 and 5 two important points become
apparent:

To avoid problems involving copper transport, the release of copper oxides must be
minimized during all modes of service, including transient conditions and idle
periods.

Copper-rich deposits formed during normal unit operation will release copper to
the cycle during startups.

Action 3, outlined in Figure 9-5, addresses the latter point; Actions 4-7 address the
former.
The first step is to collect and review available information on cycle deposits. Locations
which need to be addressed include the deaerator storage tank, boiler feed pump
strainers, feedwater heaters (waterside and steamside), boiler waterwalls, mud drums,
9-8

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

the primary superheater, turbine nozzle blocks and strainers, and the high pressure
turbine. In units with a long-standing history of copper deposition in the high pressure
turbine, it would also be desirable to assess conditions in the reheater, intermediate
pressure turbine and low pressure turbine.

Figure 9-4 Road Map for Action 2, Evaluate Condition of Copper Alloy Components
9-9

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Figure 9-5 Road Map for Action 3, Characterize Copper in Cycle Deposits

If the existence and composition of deposits in any of these areas has not been
established, arrangements should be made to do so in the future.
In view of the extent to which the volatility of copper and its solubility in steam are
influenced by the form of copper, the deposits from the boiler waterwalls and the steam
9-10

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

path should be analyzed by x-ray diffraction to identify the form of copper. Speciation
of copper should also be considered when conducting monitoring campaigns (Action 4)
and when evaluating startup procedures and layup practices (Action 5). Following this
approach provides the best opportunity to characterize copper transport, correlate
transport rates to chemistry and operating conditions (oxidizing or reducing) and, as a
result, minimize problems which result from copper-rich deposits.
In those units where initial investigative efforts indicate that cycle deposits represent a
significant inventory of copper and monitoring campaigns verify that these deposits
release copper during startups and other transient conditions, utility personnel should
then evaluate the cleaning requirements. In general terms, the criteria discussed under
Action 1A (Figure 9-2) provided good guidance on the need to clean. Final justification
to clean as soon as possible or continue to clean as required by existing criteria must
consider costs and benefits specific to the unit under evaluation.

Action 4 - Optimize Cycle Chemistry


The overall road map for cycle chemistry optimization is provided in Figure 9-6. Initial
efforts to optimize and customize chemistry should follow the guidelines given in
Selection and Optimization of Boiler Water and Feedwater Treatments for Fossil Plants (2).
Because of the overall importance of optimizing the feedwater chemistry in mixedmetallurgy units for both developing solutions and identifying problems, the nine-step
approach has been included in Section 8 (see Figure 8-1), which provides details on the
major solution options.
The feedwater treatment optimization is conducted in three phases. The first phase
involves review of current conditions including chemistry, operations and experience.
The second phase covers implementation of changes when required. The third phase
addresses verification of the effects of the implemented changes. The overall result of
the optimization should be a feedwater chemistry with less than 2 ppb copper at the
economizer inlet for normal operation; reducing conditions (ORP < 0 mV) during
operation and shutdown; and the use of an oxygen scavenger at the optimum addition
rate. It is important to also monitor the effects on the iron transport levels while
attempting to optimize copper transport.
Reference 2 also includes a road map for optimization of drum boiler water treatment.
The 20 steps included in the road map do not constrain optimization with respect to
cycle metallurgy or copper transport problems. These issues are addressed further in
the subsection entitled Conducting Monitoring Campaigns for Assessment of Copper
Transport.

9-11

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Figure 9-6 Road Map for Action 4, Optimize Cycle Chemistry

Action 5 - Improve Startup Procedures and Layup Practices


Actions 5 address minimization of copper transport when returning a unit to service.
Although startup procedures and layup practices may be evaluated separately, these
subject areas are closely interrelated. During baseline testing (Action 4), effort should
be given to monitor corrosion product transport and other cycle chemistry parameters
during the startup periods to determine the sources of copper release, including copper
9-12

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

components and deposition sites. The data collected during the optimization process
(Action 4) should be analyzed as a function of operating parameters to determine if
correlations exist between unit operation and copper transport. The monitoring
campaigns should be continued until all chemistry parameters have stabilized at values
considered typical for normal unit service. This approach must allow determination of
the duration of transport activity as well as the magnitude. The road map for Action 5
is presented in Figure 9-7.
Once corrosion product transport and other chemistry data trends at startup have been
established under baseline conditions, the optimum startup practices should be
developed. One or more of the following items relating to startup should be selected
for assessment:

Quality of water used to fill system (reduction or elimination of oxygen; chemical


treatment)

Operation of condenser air removal systems

Use of condensate polishers (where available)

Steam supply to deaerating heater (where available)

Feedwater heater startup practices

Chemical feed system activation procedures, including change in level or type of


oxygen scavenger

Routing of feedwater heater drains

Spray attemperation practices

Boiler blowdown (bottom, continuous) practices

Preparation and use of boiler water control curves for copper (See Sections 5 and 6,
and Figure 6-3)

In addressing any of these startup items, or others relating to layup included in Section
8, the primary requirement is to keep reducing conditions (ORP < 0 mV) in the
feedwater, particularly the LP systems, for as much time as possible. Emphasis should
be given to those practices which, based on monitoring results, appear to be
contributory to excessive copper transport. For example, spray attemperation practices
need not be emphasized in cases where flows are very low (or not required) at startup,
especially if the monitoring campaign indicates that most copper in the steam
originates in the boiler. Monitoring campaigns comparable to those used to define
baseline conditions during startup should be applied to assess the benefits of any
changes in startup practices adopted.
Use of copper control curves may be of some benefit in units which experience volatile
copper carryover during startups, especially in cases where copper-rich deposits exist
in the feedwater or boiler waterwalls and cleaning is either not justified or cannot be
9-13

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

performed until a future scheduled outage. However, the costs (extended startup
periods) and benefits (reduced copper transport) must be evaluated on a case by case
basis; cost-benefit aspects of deposit removal should also be factored into site-specific
considerations of developing and employing boiler water copper control curves.

Figure 9-7 Road Map for Action 5, Review Startup Procedures and Layup Practices

9-14

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

With regard to the component layup practices from the standpoint of copper transport
activity, greatest emphasis should be placed on protection of the waterside and
shellside surfaces of copper alloy feedwater heaters. However, attention should also be
given to the boiler and the turbine. Protection of the inventory of any copper on the
boiler waterwalls and in the superheaters is especially important since substantial
quantities of copper-rich deposits may be present in these areas of the cycle.

Action 6 - Examine Cycle Design


Preceding actions have considered what can be done to reduce copper transport
activity primarily through optimization of chemistry and improvement of operating
procedures. Cycle design has only been considered with respect to assessment of unit
susceptibility to copper problems, except to point out that retubing of copper-based
components with all-ferrous alloys is the only sensible choice.
Implementation of Actions 1-5 will, if properly applied, result in minimization of
copper transport activity. Further improvements may be accomplished through
examination of the cycle design and, where justifiable, retrofitting of systems capable of
improving control of copper transport activity. The road map for Action 6 is shown in
Figure 9-8. Solutions 811 in Section 8 also address the possible options.
Examples of design modification which may be considered under Action 6 include the
following.

Piping changes to enhance preboiler cleanup and/or improve oxygen removal


when placing the unit in service.

Installation of a condensate polisher or filter system sized for either unit startup or
normal operation.

Provision of a turbine bypass line for use at startup.

When considering major design changes, it is essential that all benefits to be derived
are examined. For instance, establishment of startup loops and provision of condensate
polishers will be beneficial in reducing transport of other metals besides copper. Steam
path audit findings can provide useful input to the evaluation and justification
process(3, 4). Turbine bypass systems should receive additional emphasis where
concerns exist with respect to solid particle erosion damage(5).
Enactment of a feedwater piping change to permit deoxygenation of feedwater prior to
filling the boiler enabled one utility to greatly reduce copper oxide exfoliation from the
steamside of high pressure heaters. This initiative was part of a broader program
which reduced unit startup time from more than 12 hours to less than seven hours(6).
9-15

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Action 7 - Follow-up Monitoring Campaigns


Action 7 is simply a repeat of Action 4 and of the procedures described in Section 8
under Solution 4, Optimization of Feedwater Chemistry. Its primary purpose is
obvious: to ensure the changes in feedwater chemistry, operating procedures and
maybe equipment changes and additions, produce the desired effect.
There is no room for complacency with copper alloys in the feedwater; this Action
should be repeated periodically.
SOME CONCLUDING REMARKS ON CONDUCTING MONITORING CAMPAIGNS
FOR ASSESSMENT OF COPPER TRANSPORT
For those units determined to be either currently experiencing problems related to
copper transport or at risk of such problems in the future, it is necessary to conduct
monitoring campaigns for various reasons. These include definition of baseline
conditions, identification of sources of copper release, determination of locations
subject to deposition and appraising the relative importance of available transport
paths to the main steam. Furthermore, the changes in transport activity associated with
different operating conditions need to be established. Finally, monitoring is needed to
evaluate benefits derived from any improvements made in cycle chemistry, operating
procedures, layup practices, etc.
In general, the approach to be taken when conducting monitoring campaigns is similar
to that outlined in previous EPRI publications(7-11). A comprehensive assessment of
cycle chemistry and transport must be conducted to verify that the program is working
effectively to minimize corrosion, deposition and carryover. Based on the current state
of knowledge, it is desirable to take some additional steps in units where copper
transport is already of concern (see Action 1). These issues are discussed further as
follows.

9-16

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Figure 9-8 Road Map for Action 6, Examine Cycle Design

9-17

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Sample Points
To fully characterize copper transport activity, several supplemental sample points may
need to be included in monitoring campaigns. These are outside the core requirements
of EPRI chemistry guidelines (Table 1-1) and will not be available in all units.

Low pressure heater drains

High pressure heater drains

Main steam (reheat units)

Other points at which copper should be monitored (subject to availability) include the
condensate pump discharge, condensate polisher (inlet and outlet), deaerator (inlet and
outlet), attemperation spray water, boiler feedwater (at economizer inlet), boiler water
(blowdown and/or downcomer) and reheat steam. The latter point will be important if
copper is present on the intermediate pressure turbine and/or if turbine wash
procedures are in use.

Parameters to be Monitored
All chemistry parameters indicated on cycle chemistry diagrams developed by EPRI(7-8)
may be included in the monitoring campaign. Iron testing is strongly recommended
since changes intended to reduce copper transport may result in increased transport of
iron. Chemistry parameters known to influence release and transport of copper should
also be given high priority. These include preboiler pH, dissolved oxygen and
ammonia, and scavenger levels in boiler feedwater. Condenser air inleakage rates and
condensate cation conductivity should also be evaluated.
In addition, other parameters not indicated on the EPRI guideline cycle chemistry
diagrams need to be evaluated as part of monitoring campaigns addressing copper
transport. Included here are other elements present as constituents of the copper alloys
in use (such as nickel and zinc) and feedwater oxidizing-reducing potential (ORP).
Metals analysis is useful in assessing sources of copper. Discussion in the previous
Sections have indicated that it is absolutely necessary to monitor ORP during
customization and optimization of feedwater chemistry control.

Sampling and Analysis of Corrosion Products


In most instances, use of integrated corrosion product sampler devices appears to be
the preferred means for surveillance of metals transport activity. However, collection
and analysis of grab samples may be practiced under certain circumstances. These
include:

Determination of the time required for chemistry to stabilize following return of the
unit to service.

9-18

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

Preparation and usage of unit-specific boiler water copper control curves, where
deemed necessary to reduce transport to the steam.

Assessment of metals concentrations at sample points which are not equipped with
integrated corrosion product sampler devices.

For further discussion of available sampling and analysis methods, refer to Section 5 of
this report.
For units where copper transport in the steam path is significant, consideration should
be given to use of turbine deposit collectors(12). These devices are useful for
determining the mode(s) of operation under which turbine deposition occurs and the
form in which copper exists in the steam. It may also be of some benefit in assessing
sources of copper and predicting deposition in the high pressure turbine.

Guidelines for Evaluation of Copper Transport Data


The existing EPRI chemistry guidelines (Table 2-6) for mixed metallurgy cycles include
a target value of 2 ppb for copper, measured at the economizer inlet. This value
applies to normal unit service and appears to be attainable when other cycle chemistry
parameters are maintained within guidelines. Units which are able to control copper in
accordance with the guideline have generally been found not to experience copper
problems. Units unable to achieve the copper guideline during normal operation need
to have the feedwater chemistry optimized according to the procedures outlined in
Solution 4 of Section 8 and Action 4 of this Section.
Units which are normally able to operate in accordance with EPRI chemistry guidelines
for feedwater copper may still be subject to turbine deposition and other copper
transport problems. Exceptions appear to exist in cases where one or more of the
following criteria apply.

Copper levels in feedwater during normal unit operation slightly and perhaps only
intermittently exceed 2 ppb for extended periods.

Copper transport activity is excessive only during unit startups.

Existence of either of these conditions will eventually create a substantial inventory of


copper-rich deposits that, if not removed from the cycle by appropriate cleaning
methods, will eventually be transported to the high pressure turbine. These effects
prove to be most acute in units operating at drum pressures of 2500 psig and above
(volatility enhancement), especially when the high pressure heaters feature copper
alloys (effect of temperature on corrosion and exfoliation activity).
In most units, a feedwater copper value of 2 ppb is readily attainable. In many
instances, copper levels are normally 1 ppb or less during normal operation. This
should be achievable for virtually all units if further steps are taken to optimize
9-19

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

chemistry and establish customized, unit-specific limits using the procedures and
processes provided in this report.
The more difficult issue is the degree of control over copper which should be attainable
during startup. In a semiquantitative sense, it should at least be possible to show that
the magnitude and duration of copper (and other parameter) excursions has been
reduced as a result of following the actions outlined earlier in this section with regard
to developing optimized shutdown procedures. Based on available information and
experience, the following goals for copper transport in drum boiler units are suggested
(Table 9-1).
Table 9-1
Goals for Copper in Mixed Metallurgy Drum Boiler Units
Sample Point

Maximum Copper Concentration(1)

Time to Attain Normal


Values, Hours(2)

Startup

Normal Operation

Feedwater/Economiser
Inlet

100

2 ppb

24-72

Boiler Water

500

20 ppb

8-16

1 ppb

8-16

Saturated and Main Steam


(1) Total copper, ppb Cu

(2) Time elapsed from turbine roll

These goals are considered achievable in most, but not necessarily all cases. However,
even those units which can not meet the goals but can accomplish significant
improvement in copper transport at startup (as compared to baseline conditions
determined during early monitoring campaigns) should see a significant decrease in
the magnitude of copper problems experienced.
RESOURCES FOR COPPER IN UTILITY PLANT CYCLES
The reference listings at the end of each section of this report represent a
comprehensive compilation of published resources addressing copper in fossil plant
cycles.
Other EPRI publications and products may be of interest to utility personnel
responsible for cycle chemistry that need to know more about other subjects related to
copper release from feedwater heaters. A feedwater heater performance monitor has
been developed(13) and has been used at a utility station to optimize heater
performance(14). Reports are available on other aspects of feedwater heater technology
including documents which address heater replacement(1517), leak detection(1820),
materials(2124), maintenance (25) and alternate designs (26). Condition assessment
guidelines have been developed which address feedwater heaters as well as other fossil
9-20

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

plant components(27). Significant work has been done to improve the understanding of
cycle air inleakage problems(2831). More recently, the use of sulfur hexafluoride as an
inert tracer gas to identify air inleakage sources has been evaluated and
demonstrated(32, 33). Excessive air inleakage not only aggravates corrosion and release of
copper, but reduces unit reliability, increases heat rate values and increases operating
costs.
REFERENCES
1. R.B. Dooley and W.P. McNaughton, Boiler Tube Failures: Theory and Practice. EPRI
Book TR-105261. 1996.
2. Selection and Optimization of Boiler Water and Feedwater Treatments for Fossil Plants.
Electric Power Research Institute, Palo Alto, Calif. TR-105040. 1997.
3. T. Gilchrist. Preliminary Steam Path Audit and Turbine Deposition Analysis
Results at Tri-State G&Ts Craig Station. Presented at the Fourth Annual Southwest
Chemistry Workshop, Farmington, New Mexico, July 26-27, 1995.
4. R. Devalois, T. Gilchrist and K. Price. Justification of the Retrofit of a Condensate
Polishing System at Tri-State G&Ts Craig Station. Presented at EPRI Condensate
Polishing Workshop: Deep Bed and Powdered Resin Systems. New Orleans, LA,
September 15-17, 1993.
5. K. G. Reinhard, G. D. Schatzmann and Z. S. Stys. Operating Experience with
Bypass System in Fossil and Nuclear Power Plant. Combustion, Volume 49, No. 5,
November 1977.
6. R. C. Meyer. Cold Start Procedure Improvements Cut Time and Costs and
Tombigbee Power Plant. Presented during the Operating and Maintenance Session
of the 39th Annual Conference of the Association of Rural Electric Generating
Cooperation. 1988.
7. Cycle Chemistry Guidelines for Fossil Plants: All Volatile Treatment. Electric Power
Research Institute, Palo Alto, Calif. EPRI TR-105041. April 1996.
8. Cycle Chemistry Guidelines for Fossil Plants: Phosphate Treatment for Drum Units.
Electric Power Research Institute, Palo Alto, Calif. EPRI TR-103665. Final Report,
December 1994.
9. Cycle Chemistry Guidelines for Fossil Plants: Oxygenated Treatment. Electric Power
Research Institute, Palo Alto, Calif. EPRI TR-102285. December 1994.
10. Sodium Hydroxide for Conditioning the Boiler Water of Drum-Type Boilers. Electric
Power Research Institute, Palo Alto, Calif. TR-104007. January 1995.
11. Cycle Chemistry Improvement Program . Electric Power Research Institute, Palo Alto,
Calif. TR-106371. June 1996.

9-21

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

12. O. Jonas, L. Dupee, J. Gilchrist, G. Lawrence and M. Stevens. Copper Deposition


and MW Loss: Problems and Its Solutions. Proceedings: 57th Annual International
Water Conference. Engineers Society of Western Pennsylvania. 1996.
13. Mark I Performance Monitoring Products. Electric Power Research Institute, Palo Alto,
Calif. GS/EL-5648, September 1989.
14. First Use: On-Line Feedwater Heater Performance Monitor. Electric Power Research
Institute, Palo Alto, Calif. FS 9101. January 1991.
15. Innovators with EPRI Technology: EPRI Feedwater Heater Replacement Guideline Help
NIPSCO Improve Units Heat Rate. Electric Power Research Institute, Palo Alto, Calif.
IN-102456. November 1993.
16. Feedwater Heaters - Replacement Specification Guidelines. Electric Power Research
Institute, Palo Alto, Calif. EPRI GS-6913. August 1990.
17. M. S. Rudd. The Life Cycle Economics of Feedwater Heater Replacement.
Proceedings: Feedwater Heater Technology Symposium . Electric Power Research
Institute, Palo Alto, Calif. EPRI TR-102923. September 1993.
18. Innovators with EPRI Technology: Acoustic Feedwater Heater Leak Detection Increases
Component Life, Reduces Outage Duration . Electric Power Research Institute, Palo
Alto, Calif. IN-101619. December 1992.
19. W. Woyshner, T. Bryson and M. Robertson. Acoustic Feedwater Heater Leak
Detection - Industry Application of Low and High Frequency Detection. EPRI
Fifth Predictive Maintenance Conference. September 1992.
20. Acoustic Feedwater Heater Leak Detection: Utility Experience. Electric Power Research
Institute, Palo Alto, Calif. Preliminary Report, EPRI RP-1863-4. July 1991.
21. Innovators with EPRI Technology : New Fe-Cr-Ni-Mo Austenitic Alloy Tubing Helps
Utility Improve Life of High Pressure Feedwater Heaters. Electric Power Research
Institute, Palo Alto, Calif. IN-101248. December 1992.
22. Proceedings: Feedwater Heater Technology Conference. Electric Power Research
Institute, Palo Alto, Calif. GS-7290. May 1991.
23. Corrosion - Related Failures in Feedwater Heaters. Electric Power Research Institute,
Palo Alto, Calif. CS-3184, July 1983.
24. Failure Cause Analysis - Feedwater Heaters. Electric Power Research Institute, Palo
Alto, Calif. CS-1776, April 1981.
25. Feedwater Heaters Maintenance and Repair Technology: Reducing Outage Cost. Electric
Power Research Institute, Palo Alto, Calif. GS-6935. August 1990.
26. Technical Brief: Header Feedwater Heaters Adopted in the United States. Electric Power
Research Institute, Palo Alto, Calif. TB-102286. January 1994.
27. Condition Assessment Guidelines for Fossil Fuel Power Plant Components. Electric Power
Research Institute, Palo Alto, Calif. GS-6724. March 1990.
9-22

EPRI Licensed Material


Assessment and Correction of Copper Corrosion, Transport and Deposition Problems

28. Condenser Inleakage Monitoring System Development. Electric Power Research


Institute, Palo Alto, Calif. NP-2597. September 1982.
29. Failure Cause Analysis - Condenser and Associated Systems. Electric Power Research
Institute, Palo Alto, Calif. CS-2378. June 1982.
30. Steam Plant Surface Condenser Leakage Study Update. Electric Power Research
Institute, Palo Alto, Calif. NP-2062. May 1982.
31. Location of Condenser Leaks at Steam Power Plants. Electric Power Research Institute,
Palo Alto, Calif. NP-912. October 1978.
32. Condenser Leak-Detection Guidelines Using Sulfur Hexafluoride as a Tracer Gas. Electric
Power Research Institute, Palo Alto, Calif. CS-6014. September 1988.
33. First Use: Sulfur Hexafluoride Leak Detection Helps Public Service Electric and Gas
Improve Unit Efficiency. Electric Power Research Institute, Palo Alto, Calif. FS 9022B.
May 1990.

9-23

EPRI Licensed Material

10
FUTURE RESEARCH REQUIREMENTS
Upon examination of the existing state of knowledge with respect to all aspects of
copper in fossil plant cycles, it is apparent that basic information on some subjects is
unavailable while the understanding of others needs to be improved. Table 10-1
identifies research areas which need additional activity in order to eliminate
deficiencies in the existing knowledge base.
As part of "Program Copper", EPRI has taken or is considering initiatives intended to
address some of the areas of deficiency; the current status is indicated in Table 10-2.
Further research will be sponsored pending results of these initial investigations. A
separate research program managed by EPRI in collaboration with many international
utilities has focused on improved understanding of deposition and corrosion in
turbines(1). Several EPRI publications and technical papers have been or will be issued
as a result of work performed under Phases 1 and 2 of the program(2-8). Work to date
has focused on the chemistry of early condensate, identification of deposition
mechanisms, turbine corrosion, testing devices (converging-diverging nozzles, turbine
deposit collectors, drying probes, and internal and external condensate collectors) and
molecular modeling. Eventually, it may be feasible to apply some of this information
to the mechanisms governing copper deposition in the high pressure turbine.

10-1

EPRI Licensed Material


Future Research Requirements

Table 10-1
Research Needs for Improvement of the State of Knowledge of Copper Alloys in
Utility Fossil Plant Cycles
Subject Area

Research Needs/Deficiencies to be Addressed

Corrosion of Copper Alloys

Clarify effects of pH, scavenger level, ORP, carbonates,


organics, flow and temperature.

Electrochemistry of Copper Alloys

Better define optimal reducing/oxidizing chemistry


regimes. High temperature pH-potential diagrams for
practical copper alloy materials.

Volatility of Copper

Verify/refine existing volatility data; evaluate influence


of treatment and boiler water contaminant levels on
volatility.

Solubility of Copper in Steam

Improve available data, focusing on conditions at which


copper turbine fouling is most likely.

Influence of Boiler Water Treatment

Establish correlations, if any, between specification of


copper and mechanical carryover rates. Solubility of
copper and oxides in boiler water.

Deposition of Copper in Turbines and


Superheaters

Establish mechanism(s) which govern deposition,


determine if changes in environment or materials would
reduce deposition tendencies and would be practical.

Sampling Steam and Water for Analysis Establish statements of precision and bias for current
of Copper
industry practices; determine if improvements to
existing practices are needed and practical.

Table 10-2
Initial Research to be Sponsored by EPRI under Program Copper
Subject

Status

Corrosion of Copper Alloys

Initial work addressing oxidizing/reducing potential (1997)

Volatility of Copper

Initial research study to evaluate existing information (1997).

Solubility of Copper in Steam

10-2

Work to be conducted in Denmark


(1997-1999).

EPRI Licensed Material


Future Research Requirements

REFERENCES
1. O. Jonas and B. Dooley. "Steam Chemistry and Its Effects on Turbine Deposits and
Corrosion". Proceedings: 57th International Water Conference. Engineer's Society of
Western Pennsylvania. 1996.
2. Turbine Steam Chemistry and Corrosion. Electric Power Research Institute, Palo Alto,
Calif. TR-103738. August 1994.
3. Steam, Chemistry, and Corrosion in the Phase Transition Zone of Steam Turbines. Electric
Power Research Institute, Palo Alto, Calif. TR-108184. December 1997.
4. Turbine Steam, Chemistry and Corrosion: Experimental Turbine Tests. Electric Power
Research Institute, Palo Alto, Calif. TR-108185. December 1997.
5. O. Jonas, R. Mathur and B. Dooley. "EPRI and International Projects on Turbine
Steam Chemistry and Corrosion". Presented at the Fourth International Conference
on Cycle Chemistry in Fossil Plants. TR-104502. January 1995.
6. V. Petr, M. Kolovratnik, I. Jirecek and O. Jonas. "Experimental Investigation of the
Effects of Steam Chemistry on Droplet Nucleation". Presented at the EPRI Workshop
on Moisture Nucleation in Steam Turbines . 1995.
7. M. Stastny, O. Jonas and M. Sejna. "Numerical Analysis of the Flow with
Condensation and Chemicals in Steam in a Turbine Cascade". Presented at the EPRI
Workshop on Moisture Nucleation in Steam Turbines. 1995.
8. O. Jonas. "Steam Chemistry and Condensation in Turbines". Presented at the EPRI
Workshop on Moisture Nucleation in Steam Turbines . 1995.

10-3

You might also like