Sunteți pe pagina 1din 31

1999 Tri-Service Corrosion Conference

AMPTIAC Document Number: AM026053

Environmental Effects in Flow-Assisted Corrosion of Naval Systems

Shifler, David A.

Distribution Statement:

Approved for public release and sale.


Distribution Limitation:

No additional limitations.
untitled Page 1 of 1

Attendance List
Sponsoredby
Army Research Laboratory

Organized by
Weapons Materials Research Division
Metals & Ceramics Research Branch

Conference Chairman
Dr. Ralph Adler

Editors
Mr. John V. Kelley
Mr. Brian Placzankis

To view the files,download and install the Adobe Acrobat Reader

Producedby the Document Automation & Production Service (DAPS)

file://C:\WINNT\Profiles\pmcquinn\Desktop\INDEX.HTM 3/9/04
untitled Page 1 of 1

1999 CORROSION TRI-SERVICE CONFERENCE

Monday -- November 15, 1999

Welcome Dr. John Beatty, ARL


Opening Remarks Dr. Ralph Adler, ARL
Keynote Speaker Dr. John Sedriks, ONR
"Greener, Cheaper and Better Materials for Corrosive (Marine) Environments"

Real World Issues:


"Helicopter and Ground Vehicles" CW5 Richmond Army
"AAV Corrosion Challenges" Bryan Prosser Marine Corps
"Corrosion: An Affordability and
COL Bob Tipton AF, CPO
Readiness Challenge"
"Naval Vessels" RADM Roland Knapp Navy

file://C:\WINNT\Profiles\pmcquinn\Desktop\OPEND.HTM 3/9/04
untitled Page 1 of 4

Filename Topic Authors

SESSION 1 - COATING I

S01p1a.pdf Self Healing Coating Systems and L.D. Stephenson, Tim Kang and Ashok Kumar
Accelerated Corrosion Testing
S01p1b.pdf Paul A. Decker, I. Carl Handsy, John Repp & J.
Low VOC Chip Resistant Coatings
Peter Ault, P.E.
S01p1c.pdf Robust Coatings for Corrosion and Roger N. Johnson
Wear; The Electrospark Deposition
Process
S01p2a.pdf Zero-VOC Waterbone Polyurethane Kevin J. Kovaleski
Topcoat
S01p2b.pdf A Survey of Fungal Contamination and Brenda Little, Robert Pope, Richard Ray, and
Control Measures in H-53 Aircraft Lynne Pfledderer
S01p2c.pdf Effects of Nitrogen on the J. D. Demaree, W. E. Kosik, C. R. Clayton and
Electrochemical Passivation of Metal G. P. Halada
Nitride Coatings Produced by ION
Beam Assisted Deposition

SESSION 2 - ENVIRONMENTAL ISSUES

S02p1a.pdf Evaluation Aluminum-Manganese as a Dr. Michael J. Kane, Craig Matzdorf, and Jim
Cadmium Replacement Green
S02p1b.pdf Novel Electrochemical and K. Rajeshwar
Photochemical Approaches to
Hexavalent Chrome and Lead
Reduction
S02p1c.pdf Corrosion Testing for Alternative Wayne Ziegler, William Newton, Albert J.
Solvent Substitution Performance Walker, Jr.
Validation
S02p2a.pdf Paul F. Buckley, Brian E. Placzankis,
Corrosion of Buried Ordnance:
Christopher E. Miller, Thomas F. Curry, James
Fundamentals, Testing and Initial
D. Garber, Cedric D. Adams, and Joseph E.
Modeling
Bucci
S02p2b.pdf Corrosion Behavior of Aluminum M. Pujar, C. Arvin, G. Banerjee, K. L. Vasanth,
Alloys Treated with New Green and A. E. Miller
Inhibitors
S02p2c.pdf Subsitution Pollution Prevention William B. Campbell and Tiki Dixon

SESSION 3 - CORROSION MONITORING/SENSOR

S03p1a.pdf Investigation of Corrosion Initiation A. Buchanan, M. Khobaib, T. Matikas, and M.


Using the Scanning Vibrating Donley
Electrode Technique
S03p1b.pdf In-Situ Sensor to Detect Moisture G.D. Davis, C.M. Dacres, and L.A. Krebs
Intrusion and Degradation of Coatings,
Composites, and Adhesive Bonds
S03p1c.pdf Corrosion Database Management Carl Handsy and Dr. Tom Savell
System
S03p1d.pdf In-situ Measurement of Corrosion N.K. Batra, K.E. Simmonds and R.B. Mignogna
Damage in Storage Tanks Containing
Liquids Using Ultrasonic Spectral

file://C:\WINNT\Profiles\pmcquinn\Desktop\PAPERS.HTM 3/9/04
untitled Page 2 of 4

Tracking
S03p2a.pdf Corrosion Monitoring of Air Force Field W.H. Abbott
Sites
S03p2b.pdf A Tale of Two CHT Tanks R.L. Ruedisueli, R.A. Hays, and J.N. Murray
S03p2c.pdf Methods for Detection of Corrosion in Mr Iain McKenzie and Dr. Peter Newman
Aircraft Structures Using Fibre Optic
Technology

SESSION 4 - COATINGS II

S04p1a.pdf Evaluation of Coatings to Reduce A.D. Sheetz and E.B. Bieberich


Corrosion in Crevices for Marine
Corps Vehicles
S04p1b.pdf Comparison of Laboratory Corrosion J. F. Dante, L. Petry, and D. T. Peeler
Test Methods and Outdoor Corrosion
Tests for Air Force Coating Systems
S04p1c.pdf Corrosion Controls of Steels by Silane Dr. W.J. van Ooij and Yuan Fu
Coatings
S04p2a.pdf Sol-gel Coatings with Organic Stuart F. Cogan, Michael D. Gilbert, Peter J.
Corrosion Inhibitors Marren, Julia Ehrlick, Joseph H. Osborne and
Dennis I. Dull
S04p2b.pdf Poly(acid)-modified Chitosan Marine T. Sugama
Polymer Coatings for Aluminum
Substrates
S04p2c.pdf AN/SLQ-32 Antenna Enclosure R. K. Conrad, C. H. Cox 1 , C. Bowles, and
Coating Investigation and EMI/EMP D.E. Hileman
Improvements

SESSION 5 - ACCELERATED TESTING

S05p1a.pdf Accelerated Corrosion Testing in a I. Carl Handsy and J. Peter Ault, P.E.
Static Marine Environment
S05p1b.pdf Corrosion Durability Testing of Military I. Carl Handsy, Paul A. Decker, J. Peter Ault,
Vehicles P.E. and John P. Repp
S05p1c.pdf Accelerated Corrosion Test Facility Steven E. King
S05p1d.pdf Comparison of Accelerated Test E.B. Bieberich, R.A. Hays, and A.D. Sheetz
Results to Marine Atmospheric
Exposure for U.S. Marine Corps
Applications
S05p2a.pdf Accelerated Corrosion Analysis of E- Brian E. Placzankis, Chris E. Miller, Paul F.
Coated Steel Box Panels to Determine Buckley and John H. Beatty
Reapplication Intervals for Carwell
Inhibitor
S05p2b.pdf Accelerated Corrosion Analysis of Brian E. Placzankis, Chris E. Miller and John H.
Nonchromate Conversion Coatings on Beatty
Aluminum Alloys 5083, 7039, and
6061 for DoD Applications
S05p2c.pdf Comparison of Water Reducible Chris E. Miller and Brian Placzankis
CARC versus Standard CARC on
Steel
Using Three Accelerated Test
Methods

SESSION 6 - ECONOMICS/LIFECYCLE ISSUES

S06p1a.pdf Corrosion Management - A Key Vinod S. Agarwala


Challenge for Reduction of Total
Ownership Cost
S06p1b.pdf Corrosion Wording for USMC R.A. Hays and E.B. Bieberich

file://C:\WINNT\Profiles\pmcquinn\Desktop\PAPERS.HTM 3/9/04
untitled Page 3 of 4

Acquisition Documents
S06p1c.pdf USCG Aviation Corrosion Control
Program
S06p2a.pdf NAVAIR Perspective on Corrosion Stephen J. Spadafora
Prevention and Control
S06p2b.pdf Application of Practical Aging R.M. Latanision, G.G. Leisk, W.J. McBrine,
Management Concepts to Corrosion T.C. Esselman, O.J. Van Der Schijff and P.J.
Engineering Wender

SESSION 7 - CORROSION ENGINEERING

S07p1a.pdf Denise M. Aylor, Richard A. Hays, and Robert


Corrosion-Resistant Seawater Valves
J. Ferrara
S07p1b.pdf The Advantages of Nickel Alloys for Edward L. Hibner and Lewis E. Shoemaker
Seawater Service
S07p1b.pdf Laboratory and Field Evaluations of P. Trzaskoma-Paulette, S. G. Lambrakos and
Rigid Polyurethane Foams as a Harry N. Jones III
Retrofit Method of Corrosion Control
for Military Equipment
S07p2a.pdf Corrosion Testing in Support of the R.A. Hays, E.B. Bieberich, A.D. Sheetz, and J.
Advanced Amphibious Assault Vehicle Chrzan
S07p2b.pdf Corrosion Evaluation of Washdown E.B. Bieberich and A.D. Sheetz
Additives for Marine Corps Vehicles
S07p2c.pdf Possible Control Measures for David A. Shifler
Galvanic Corrosion in Piping Systems

SESSION 8 - ENVIRONMENTALLY ASSISTED


CRACKING

S08p1a.pdf Stress Corrosion Cracking D. A. Davis, E. J. Czyryca and J. A. Manuel


Susceptibility of the Heat-Affected
Zone of High Heat Input HY-80 & HY-
100 Steel Weldments
S08p1b.pdf Fatigue Characterization of Crack M. Khobaib, T. Matikas and M.S. Donley
Nucleating from Corrosion Pits
S08p1c.pdf Effect of Diffusible Hydrogen Richard L.S. Thomas, Richard P. Gangloff, and
Concentration on Internal Hydrogen John R. Scully
Embrittlement of AerMet 100
S08p2a.pdf Stress Corrosion Cracking of High Eun U. Lee, Henry Sanders, and Bhaskar
Strength Steels Sarkar
S08p2b.pdf Analysis of Hydrogen Trapping in Paul F. Buckley, Jeffrey A. Fagan and Peter C.
Palladium via Modulated Permeation Searson
Spectroscopy

SESSION 9 - CORROSION ENGINEERING II

S09p1a.pdf Chloride Contamination Leads to TI- Wayne Ziegler, Victor K. Champagne, Scott
6AL-4V Blackhawk Compressor Grendahl, and Marc Pepi
Impeller Failures
S09p1b.pdf Environmental Effects in Flow- David A. Shifler, PhD, PE
Assisted Corrosion of Naval Systems
S09p1c.pdf Effects of Fuel Carbon Residue on the David A. Shifler
Hot Corrosion Resistance of Materials
in Marine Gas Turbines
S09p2a.pdf Corrosion in Wire Rope Spelter James F. Lewis
Sockets
S09p2b.pdf Corrosion Testing of Turbine Engine G.Y. Richardson, D. L. Stone, and J. N. Carroll
Seal Materials

file://C:\WINNT\Profiles\pmcquinn\Desktop\PAPERS.HTM 3/9/04
untitled Page 4 of 4

SESSION 10 - CORROSION RESEARCH

S10p1a.pdf The Hot Corrosion Resistance of High- David A. Shifler and Leslie K. Kohler
Chromium Nickel-Based Alloys
Fabricated by Sprayed Forming
S10p1b.pdf Intergranular Corrosion of High W. Zhang, T. Ramgopal and G. S. Frankel
Strength Al Alloys
S10p1c.pdf An Electrochemical Impedance Jack V. Kelley, Costas G. Fountzoulas and
Evaluation and Laser Irradiation Garry P. Halada
Effects on the Electronic Structure of
Silicon Containing Diamond-Like
Carbon Coating
S10p1c.pdf Modeling and Quantification of W. M. Mullins and R. L. Crane
Corroded Surfaces
S10p2c.pdf Spatially-Resolved Microchemical Analysis Gary P. Halada, Clive R. Clayton, Marvin J.
of Chromate Conversion Coated Aluminum Vasquez, Jeffery R. Kearns, Martin W. Kendig,
Alloys and Constituent Intermetallic Samuel L. Jeanjaquet, Gregory G. Peterson,
Particle Analogs Grace Shea McCarthy, G. Lawrence Carr,
Gwyn P. Williams, and Lisa M. Miller

SESSION 11 - SURFACE TREATMENTS

S11p1a.pdf Maximizing Coating Adhesion Through Christine A. Bowles and Dr. David A. Shifler
the
Removal of Surface Contaminants
S11p1b.pdf Organic Corrosion Inhibitors for Al M. Khobaib, Natalia Voevodin and M. S.
2024-T3 Donley
S11p2a.pdf Demonstration and Validation of Craig A. Matzdorf, Dr. Michael J. Kane, James
Trivalent Chromium Aluminum L. Green, and Tim Woods and Mike Seybold,
Pretreatment on U.S. Navy S-3 Aircraft
S11p2b.pdf Non Chrome Sealers for Aluminum Robert W. Katz and Jules F. Senske
Anodizing
S11p2c.pdf Effects of Vapor Phase Corrosion K. L. Vasanth, Seok-Hwan Chung and R. D.
Inhibitor (VCI) Emitters on Magnetic Gomez
Storage Devices

file://C:\WINNT\Profiles\pmcquinn\Desktop\PAPERS.HTM 3/9/04
The Effect of Acetone Cleaning on the Behavior of AA2024-T3 in a Mist of 0.5 M NaCl.
Improving Aluminum Ion Vapor Deposition
Depleted Uranium Predictive Technologies: Accelerated Corrosion Testing
Surface Pretreatments for Chromate Conversion Coating of AA2024-T3 Alloy
Heterogeneity in Commercial Coating Systems for Aerospace Applications
A Study of the Interaction of Contaminant Uranium with Atmospheric Corrosion Products on Steel
and Subsequent Complexation by Hydroxycarboxylic Acid
TITANIUM ALLOY 5111: Brings Intermediate Strength, Excellent Toughness, and Corrosion
Resistance to Naval Operating Environments
ENVIRONMENTAL EFFECTS IN FLOW-ASSISTED
CORROSION OF NAVAL SYSTEMS

David A. Shifler, PhD, PE


Carderock Division
Naval Surface Warfare Center
Code 613, Corrosion Branch
9500 MacArthur Boulevard
West Bethesda, MD 20817-5700

ABSTRACT

Flow-assisted corrosion can take many forms: (1) mass transport controlled corrosion, (2) phase transport
controlled corrosion, (3) erosion corrosion, and (4) cavitation. Flow-assisted corrosion is dependent on a
number of factors such as alloy composition, water chemistry, pH, biofouling, microbiologically influenced
corrosion, pitting and crevice corrosion, water pollution and contamination, alloy surface films, geometry and
surface roughness, galvanic interaction, fluid velocity and mode, oxygen content, heat transfer rate, and
temperature. Several case studies will be reviewed and examined in terms of the contributing factors to flow-
assisted corrosion in naval based systems.

Keywords: Flow-induced corrosion, flow-influenced corrosion, flow-assisted corrosion, flow-accelerated


corrosion, cavitation, erosion corrosion, boiler corrosion, carbon steel, copper-nickel, copper alloys

INTRODUCTION
Corrosion rates and the type of corrosion are often dependent on environmental factors such as fluid flow
and the availability of appropriate species required to drive electrochemical reactions [1]. A change in the
motion of a corroding metal or alloy relative to its environment by fluid flow can increase corrosion rates by
removing protective films or by increasing the diffusion or migration of deleterious species. An increase in fluid
flow can decrease corrosion rates by eliminating aggressive ion concentration or enhancing passivation or
inhibition by transporting the protective species to the fluid/metal interface. Accelerated corrosion due to flow
has been termed flow-assisted, flow-influenced, flow-induced, or flow-accelerated corrosion. There are several
mechanisms described by the conjoint action of flow and corrosion that result in flow-influenced corrosion: (1)
Mass transport-controlled corrosion; (2) Phase transport controlled corrosion; (3) Erosion corrosion; and (4)
Cavitation [2].
Mass transport-controlled corrosion implies that the rate of corrosion is dependent on the convective mass
transfer processes at the metal/fluid interface. When steel is exposed to oxygenated water, the initial corrosion
rate will be closely related to the convective flux of dissolved oxygen towards the surface, and later by the
oxygen diffusion through the iron oxide layer [3]. Corrosion by mass transport will often be streamlined and
smooth.
Phase transport-controlled corrosion suggests that the wetting of the metal surface by a corrosive phase is
flow dependent. This may occur because one liquid phase separates from another or because a second phase
forms from a liquid. An example of the second mechanism is the formation of discrete bubbles or a vapor phase
from boiler water in horizontal or inclined tubes in high heat-flux areas under low flow conditions (departure
from nucleate boiling). The distribution and morphology of the corrosive attack will be related to the distribution
of the corrosive phase; the corroded sites will frequently display rough, irregular surfaces and be coated with or
contain thick, porous corrosion deposits.
Erosion corrosion is associated with a flow-induced mechanical removal of the protective surface film that
results in subsequent corrosion rate increases via either electrochemical or chemical processes. It is often
accepted that a critical fluid velocity must be exceeded for a given material. The mechanical damage by the
impacting fluid imposes disruptive shear stresses or pressure variations on the material surface and/or the
protective surface film. Erosion corrosion may be enhanced by particles (solids or gas bubbles) and impacted
by multi-phase flows. Increased flow stream velocities and increases of particle size, sharpness, density, and
concentration increase the erosion corrosion rate. Increases in fluid viscosity, density, target material hardness,
and/or pipe diameter tend to decrease the corrosion rate. The morphology of surfaces affected by erosion
corrosion may be in the form of shallow pits or horseshoes or other local phenomena related to the flow
direction.
Cavitation sometimes is considered a special case of erosion corrosion and is caused by the formation [4]
and collapse of vapor bubbles in a liquid near a metal surface. The basic mechanisms of boiling and cavitation
are similar but differ in their thermodynamic paths towards vaporization [5]. Boiling causes rupture of a liquid
by increasing the temperature at approximately constant pressure. Cavitation is a process by which a liquid is
ruptured by decreasing the pressure at roughly constant liquid temperature. Cavitation removes protective
surface scales by the implosion of gas bubbles in a fluid; calculations have shown that the implosions produce
shock waves with pressures approaching 60 ksi [6]. Attack is the result of hydro-mechanical effects from
liquids in regions of low pressure that have experienced flow velocity changes, disruptions, or alterations in flow
direction. The time between nucleation and collapse of cavitating bubbles is usually only a few thousandths of a
second. The rate of erosion due to cavitation is dependent on bubble size and bubble impact velocity on the
alloy surface [5]. The level of the cavitation damage event does not need to be extensive. A very small bubble
fraction actually causes cavitation damage [7]. Cavitation damage often appears as a collection of closely-
spaced, sharp-edged pits or craters on the surface.
Flow patterns through pipes and tubing are governed by the laws of hydrodynamics. Hydrodynamic
factors that contribute to the rate of erosion corrosion include velocity, the surface shear stress, the turbulence
intensity, and the mass transfer coefficient [8]. The velocity field can be described by: (1) the Navier-Stokes
equation that is based on the conservation of momentum, and (2) the continuity equation which is based on the
conservation of mass. Laminar flow occurs for Reynolds numbers (Re) of about 2000 or less. Turbulent flow is
generally acknowledged for fluid Reynolds numbers exceeding 10,000. Flows with Reynolds numbers between
2000 and 10,000 may be laminar or turbulent and should be avoided because of the uncertainty of predicting
erosion corrosion rates of materials in this fluid regime. Pressure drops due to friction differ markedly in the
laminar and turbulent regimes. The Reynolds number is defined as Re = DV/ where D is the hydraulic
diameter, V is the fluid flow velocity, is the fluid density, and is the absolute or dynamic fluid viscosity.

2
Flow-accelerated corrosion (FAC) has been a persistent problem in the power industry. Factors such as
velocity, void fraction and steam quality, geometry of the component, water chemistry, alloy content and
temperature all influence the flow-accelerated corrosion rate in power plant systems [9].
It has been suggested [10] that the corrosion rate by cavitation can be influenced by surface roughness that
permits a large number of nuclei for bubble formation. Under this condition, the concentration of deleterious ions
in solution next to the metal surface will be greater, and the observed corrosion damage indicates that the steam
bubbles may provide crevices or enhanced possibilities for dissolution at the solution/metal/steam interface.
When activation processes involving electrochemical, electron-transfer reactions at the metal-electrolyte
interface were rate-controlling, small temperature rises above the equilibrium boiling point increased dissolution
by up to two orders of magnitude [11]. In boilers operating with sliding pressure loads, a sudden reduction of
pressure can release steam bubbles in lower boiler sections [12]. The associated steam release in these sections
causes localized erosion corrosion, particularly along the top interior surfaces of horizontal and inclined tubes.
Kim discusses water hammer (sudden pressure waves created by a rapid change in flow velocities in a pipe line)
in single and two-phase situations [13]. While single-phase water hammer due to sudden opening or closing of
valves in a piping system is more common, two-phase water hammer is more subtle and complex. Two-phase
water hammer may involve vapor bubble collapse, vapor column rejoining, or liquid slug impact on vapor.
Erosion corrosion of carbon steel and low alloy steels has been reported in steam-water piping under single and
dual phase flow conditions up to 300 C [14].
Although it is generally considered that flow-assisted corrosion is related to excessive fluid velocities, flow-
assisted corrosion can be caused by low flow or stagnant conditions as well. As one example, the design of all
boilers requires a sufficient fluid flow through heated rising tubes to balance heat transfer through nucleate boiling
and maintain adequate tube-metal temperatures. However, if coolant flow is inadequate to balance heat
transfer, the volume of steam becomes too great to maintain individual bubble formation. A continuous steam
film, referred to as departure from nucleate boiling (DNB), develops and tube metal temperatures rise rapidly.
Dissolved or suspended solids and chemical concentrate under these conditions, dissolving the magnetite oxide
film on carbon steels and forming thick, porous waterside deposits [15]. Low flow conditions also can allow
deposits, microorganisms, and chemicals to settle on susceptible alloy surfaces which subsequently lead to
pitting and crevice corrosion.

FACTORS AFFECTING EROSION CORROSION

Natural Waters and Process Water

Boiler water treatment attempts to prevent scale formation and corrosion of boilers and their components.
All natural waters contain varying amounts of dissolved and suspended matter. Settlement tanks and filtration
remove suspended matter. Various filtration, purification, dimineralizer, distillation condensate polishers remove
water impurities, oxygen ( 7 ppb O2), and other gases in many boilers to develop a reducing environment. A
thin magnetite, Fe3O4, oxide forms by the direct reaction of water and carbon steel in this reducing environment.
Chemical treatments are usually added to control corrosion and scale formation. Oxygenated treatment (50-
150 ppb O2) [16] provides an oxidizing environment and forms a network of magnetite and hematite, Fe2O3
[17]. The pH is maintained between 8.0 and 9.6. In both oxidizing and reducing environments, air-inleakage is
undesirable by causing corrosion rates of boiler components. Lower oxygen content, lower pH, and the
presence of a two-phase medium of steam and water reportedly increase FAC in boiler environments [18].

3
Seawater is an aggressive, complex fluid which affects nearly all common structural materials to some
extent [19]. There are two competing processes that operate simultaneously in seawater environments: (1) the
chloride ion activity towards destroying the passive film, and (2) dissolved oxygen, acting to promote and repair
the passive film on metallic materials of construction [20]. Hoar states that passivation is delayed or destroyed
by chloride ions in solution for metals and alloys that passivate by the formation of thin metal oxides [21]. The
pH in open seawater varies from 7.4 to 8.4 and is buffered by a complex carbonate system. Temperature,
dissolved oxygen, and salinity vary with geographical location, ocean currents, weather, and changing seasons,
though the ratio of concentration of major constituents is relatively constant. Corrosivity of different alloys has
been monitored over a number of years. Salinity is less important than temperature and dissolved oxygen. The
corrosivity of metallic materials in dependent on the degree of fouling, silting, bacteria activity, temperature,
dissolved oxygen, flow rate, and pH [22].

Microbiological Corrosion

Biofouling and microbiologically influenced corrosion (MIC) are possible in neutral waters, pH 4-9 and at
temperatures ranging from -10C to 99C [23], in contact with nearly all metallic materials of construction. The
formation and adherence of a biofilm usually occurs within a short time after exposure to the aqueous
environment. Biofilms then affect the interaction between the metal surface and the environment. Growth of
microorganisms requires nutrients that metal surfaces often adsorb and can form synergistic communities with
other microorganisms, plants, and animals. MIC can influence corrosion by forming surface deposits which
cause under-deposit corrosion through the development of concentration cells or by causing small flow
perturbations. MIC can cause metal losses by a variety of processes in both anaerobic and aerobic conditions.
Sulfate reducing bacteria (SRB) causes corrosion of iron, mild steels, and stainless steels in anaerobic
conditions by the formation of sulfides while sulfide-oxidizing and iron-oxidizing cause damage under aerobic
conditions. Copper alloys are more resistant because of the natural toxicity of its corrosion products, but
certain strains of bacteria can attack copper alloys, particularly those containing iron and manganese and under
microbial colonies and tubercules. SRBs and the decomposition of biofilms and microbial colonies during
which H2S, CO2, and NH3 are produced can cause corrosion of copper alloys[23]. Only titanium and some
Ni-Cr-Mo alloys have shown resistance to MIC. Other references provide details of general MIC [24,25] and
as related to cooling water systems [26]. Biofilms also cause ennoblement of stainless steels [27], alloy 625 and
titanium by increasing the cathodic kinetics, leading to a shift of the cathodic polarization curves to higher
currents and resulting in increased galvanic corrosion [28,29].
Macrofouling by barnacles, mollocks, mussels, clams, and similar organisms foul and block condensers
inlets, water supply systems, and cooling water piping and other cooling water systems. This can create flow
disturbances and generate turbulent and localized high-flow conditions. Figure 1 shows macrofouling of a
titanium pipe after six months operation in strained, flowing seawater.
General Corrosion Processes

Copper-nickel alloys are susceptible to accelerated corrosion in seawater containing high sulfide, low pH
concentrations, and high total organics [30]. Hack and Gudas [31] showed that 90/10 Cu-Ni was susceptible
to sulfide induced pitting in natural, flowing seawater containing 0.01 ppm or more sulfides. 70/30 Cu-Ni was
similarly susceptible to sulfide-induced pitting, but required higher sulfide concentrations. Syrett et al. [32,33]
observed that the presence of sulfides interfered with the formation of normal passivating films of copper nickel

4
alloys found in unpolluted waters. A black porous, nonprotective cuprous sulfide film was formed that reduced
the charge transfer resistance. Sulfide modified films generally were more loosely adherent than the normal
cuprous oxide films; turbulence tended to selectively remove the sulfide-modified films. The exposed surfaces
are anodic to the surfaces covered by the films and localized corrosion at the exposed sites is promoted.
Depending on water quality, additional sulfide-modified films may form and the process becomes cyclic. The
accelerated corrosion rates of copper-nickel alloys in aerated sulfide-containing seawater remain high since the
sulfides prevent protective corrosion product layers from forming [34].
The measured corrosion potentials of 90/10 and 70/30 copper-nickel alloys with sulfide modified films
shifted to more noble (electropositive) values than copper-nickel with the normal Cu2(OH)3Cl and Cu2O
corrosion product films [31, 35]. The magnitude of the noble potential shift tended to increase, in aerated
flowing seawater, with increasing sulfide content or longer exposures times [31]. This electropositive potential
shift was attributed to breakdown of the normal, protective films and pitting by local galvanic interaction
between the sulfide-modified films and freshly exposed copper-nickel surfaces. The ease by which sulfide-
modified films are removed and the measured potential differences support the electrochemical/mechanical
nature of pitting of copper-nickel alloys [31]. Later, it was considered that the electropositive shifts caused the
sulfide modified Cu2O film to break down locally and promote pitting at these sites.
Sulfides can be generated in several ways: (1) bacterial reduction of naturally occurring sulfates in seawater;
(2) rotting vegetation; and (3) industrial waste discharge. Exposure of new 90/10 Cu-Ni samples to putrid
seawater for 24 hours produced pitting up to 0.007-in. deep and a corrosion rate of 0.038 in./yr. [36].
Lack of flow either can exasperate crevice corrosion which takes place with passive oxide film formers
such as carbon steel and stainless steels where occluded, stagnant areas become oxygen-deficient anodic sites
as compared to outer cathodic, oxygenated regions [37,38]. A crevice gap must be wide enough for liquid to
enter, but sufficiently narrow to maintain a stagnant zone. Elevated temperatures, oxygen concentrations, and
accelerated oxygen transport to the relatively large cathodic areas outside the crevice augment the crevice
corrosion rate [39]. The corrosion rate at pitted sites also can be enhanced by similar factors. Concentration
cells are one of the major causes for pitting and is the driving force behind crevice corrosion.
Under-deposit attack or poultice corrosion may occur when a metal is locally covered by foreign,
absorbent (organic or inorganic) materials [40,41]. In this case, attack can proceed even when the bulk of the
system is dry due to retention of moisture in the poultice. The corrosion mechanism is similar to crevice
corrosion in that the deposits act to limit the migration of oxygen to the covered area. This leads to acidic shifts
in pH, concentration of Cl- ions in the shielded area, and a shift to a more active corrosion potential.
Accelerated corrosion is caused by chemical attack and/or by galvanic interactions between the relatively noble
area outside the deposit and the active area under the deposit. Local corrosion rates can be very high due to
the large cathode-to-anode area ratio. Under-deposit attack has been observed in 70/30 Cu-Ni at flow
velocities below 3 ft/sec. This type of attack has also been shown to be possible under deposits and scales
containing sulfides [42].
Alloy Surface Films

The nature and properties of surface films that form on many metals and alloys are extremely important
towards resistance to erosion corrosion. Most alloys form oxide or oxyhydroxide films in aqueous solutions.
The stability and solubility of the oxide is dependent on the chemical composition of the alloy. Chromium has

5
proven to be most beneficial towards improving the properties of the passive film of ferrous and nickel-based
alloys. Molybdenum (Mo) added to the alloys improves the pitting resistance; some of the most corrosion
resistant Ni-Cr-Mo alloys possess in excess of 15 wt. percent Mo. Titanium forms a tenacious TiO 2 oxide film
which is resistant in most oxidizing and reducing environments. Titanium and some nickel-chromium-
molybdenum alloys perform well in low, intermediate, and high flow velocities [20]. Oxide passive films that
contain insufficient molybdenum, such as in many nickel-based alloys and stainless steels, are susceptible to
pitting in stagnant and low flowing seawater, but perform well on boldly exposed surfaces at intermediate and
high flow velocities. Small alloy additions can have a marked influence on FAC resistance. Carbon steels are
susceptible to FAC, but low alloy chromium steels such as 1 % Cr- % Mo and 2 % Cr-1% Mo (ASME
SA213 T-11 and T-22 steels respectively) are very resistant to FAC.
Unlike the passive film formers, the corrosion product layers formed on copper-nickel alloys exposed to
seawater depend on the alloy and water composition. In unpolluted seawater, a loosely adherent porous cupric
hydroxy-chloride (Cu2(OH)3Cl) corrosion product scale forms over a thin, tightly adherent layer of cuprous
oxide (Cu2O) that increases corrosion resistance with increased exposure times [43,44]. The iron and nickel
from the copper-nickel alloy are concentrated in the inner portion of the porous layer [45]. Iron, if present in
solid solution, also increases the corrosion resistance of copper-nickel alloys by assisting in the formation of the
protective film [46]. The initial corrosion rate of copper-nickel alloys is relatively high but quickly decreases as
the oxide film and the corrosion product layers are formed. North and Pryor [47] observed that the cuprous
oxide film retarded both the anodic and cathodic reactions, thereby reducing the corrosion rate of copper-nickel
alloys. About two weeks are required to passivate the copper-nickel surfaces in flowing, unpolluted seawater.
The addition of another element to an alloy may increase its resistance to erosion corrosion. The effect of
iron content on the corrosion and impingement resistance of 90/10 copper-nickel is maximized with the addition
of about 2% iron to the cupronickel alloy [48]. Solutionized iron is beneficial and responsible for increased
erosion corrosion resistance through the incorporation of iron oxides into the corrosion product film [49].
Nickel and iron incorporate into the corrosion product film in flowing seawater [44,49]. Efird [50] concluded
that iron additions were necessary for nickel to enrich the Cu2O corrosion product layer and that iron needed to
be solutionized to be effective. If iron was precipitated in cupronickel alloys to form iron and nickel-iron second
phases, deterioration of the passive film occurred which altered the corrosion behavior of the cupronickel alloy
[46,50-52]. The addition of 0.5 wt. % chromium vastly improves the nature of the passive film of copper nickel
alloy UNS C72200. The iron and chromium in C72200 must be kept in solution to avoid pitting corrosion. The
critical velocity of C72200 at 27 C is 12.0 m/s (39.4 ft/s) [53]; the amount of chromium in solid solution is
important.
Copper alloys subjected to sulfide (as mentioned previously), carbon dioxide, or complexing chemicals
such as ammonia are subject to modification of the passive films that subsequently lead to breakdown of
passivity and increased metal losses.

Galvanic Interactions
The coupling of dissimilar alloys in conducting, corrosive solutions such as seawater can lead to accelerated
corrosion of the more anodic, electronegative alloy and protection of the more cathodic, electropositive alloy by
galvanic corrosion. The extent of galvanic corrosion depends on factors such as: (1) the effective area ratio
between the anodic and cathodic members of the couple; (2) solution conductivity, (3) flow characteristics of

6
the solution; (4) temperature; (5) system geometry; (6) the potential difference of the dissimilar alloys; (7)
solution composition and environment and (8) the cathodic efficiency of the more noble metal or alloy [54-56].
Use of dissimilar pipe alloys for cooling system piping or condenser tubing materials different from the tube
sheet/waterbox have the potential to cause corrosion at each change-of-material interface. Galvanic isolation
and electrically isolated separator pieces, or coating the anodic materials may not always be viable control
methods due to the complexity or available space of the systems. Alloy 20 stainless steels are more
electropositive than copper alloys such as copper nickels, brasses, and bronzes [57]. Galvanic corrosion and
erosion corrosion can attack synergistically the passive films of susceptible alloys. Galvanic action between
dissimilar metals can destroy the passive film of the more anodic member exposing the bared substrate to
possible erosion corrosion. The removal of the protective films on copper-nickel alloys by flow may promote
dissolution of the bare, exposed areas by galvanic interaction with the surrounding surfaces covered by
corrosion product scales. Local hot spots may induce thermogalvanic effects between two sites of different
temperatures for the same [11].
The effective galvanic length in piping systems is the distance that galvanic current flows due to the presence
of a dissimilar metal couple. The effective galvanic interaction lengths are dependent on factors such as the type
of alloys involved in the galvanic couple, degree of polarization of the different alloys, salinity and conductivity of
the seawater, and seawater temperature. Such measured interaction lengths range from 15 to over 200 pipe
diameters [58-61]. Reductions of galvanic current may occur in aerated seawater. Steel coupled to stainless
steels and copper-nickel will have the tendency to form calcareous deposits on the cathodic members of the
couple, thereby reducing galvanic currents.

Fluid Movement

Fluid movement, usually characterized by velocity, can have an extreme effect on the corrosion behavior of
metals and alloys. Materials must endure stagnant conditions during times when a ship is in port or between
operation cycles, or when a boiler is in a planned or unscheduled down time. Systems may utilize process
water, freshwater, or seawater and may operate in intermediate velocity regimes or high velocities approaching
the critical shear stresses for system components. In any given velocity domain, different local velocities may
exist over diverse areas of the component due to factors such as geometry or mode of fabrication [20].
Increased fluid velocities can enhance the rapid transport of aggressive anions and species such as oxygen which
can contribute to changes in the corrosion and erosion corrosion rates. Turbulent flow increases agitation of
waters with the metallic component or structure more than in laminar flow, with the probability of increased
erosion corrosion.
Titanium and some nickel-chromium-molybdenum alloys perform well in low, intermediate, and high
velocity ranges up to 120 fps (36 m/s). Oxide passive film formers such as nickel-based alloys and stainless
steels are susceptible to pitting in stagnant and low flowing seawater, but perform well on boldly exposed
surfaces at intermediate and high form velocities. They are subject to crevice corrosion. On the other hand,
copper and copper alloys perform well at low velocities in unpolluted seawater, but experience excessive
erosion corrosion losses at intermediate and high velocities. Commercial 90-10 copper-nickel is generally
considered to have acceptable performance up to 12 fps (3.6 m/s) while 70/30 copper-nickel has an upper
velocity limit of 15 fps (4.5 m/s). Other copper alloys have lower critical velocities than copper nickels.

7
Severe pitting of 90/10 copper-nickel piping caused by excessive, turbulent flow was observed downstream of
an orifice plate. This turbulent flow caused erosion-corrosion or impingement attack related pitting.
Surface shear stress is a measure of the force applied by fluid flow to the corrosion product film. For
seawater, this takes into account changes in seawater density and kinematic viscosity with temperature and
salinity [62]. Accelerated corrosion of copper-based alloys under velocity conditions occurs when the shear
surface stress exceeds the binding force of the corrosion product film. Alloying elements such as chromium
improves the adherence of the corrosion product film on copper alloys in seawater based on measurements of
the surface shear stress. The critical shear stress for C72200 (297 N/m2, 6.2 lb f/ft2) far exceeds the critical
shear stresses of both C70600 (43 N/m2, 0.9 lb f/ft2) and C71500 (48 N/m2, 1.0 lb f/ft2) copper-nickel alloys
[53, 62].
Copper-based alloys are susceptible to a critical surface shear stress in seawater where the corrosion
product film begins to breakdown and is removed from the alloy surface, subsequently causing accelerated
attack [53]. The generally accepted critical velocity for 70/30 copper-nickel in unpolluted seawater is about 15
fps (4.5 m/s) [53,63]. Copper-based alloys are susceptible to a critical surface shear stress in seawater where
the corrosion product film begins to breakdown and removed from the alloy surface thereby initiating
accelerated attack [53]. The generally accepted critical velocity for 90/10 copper-nickel (C 70600) is about
12 fps (3.6 m/s) [53,63] for large piping sizes before erosion corrosion becomes a problem. This corresponds
to a critical shear stress of 43 N/m2 (0.9 lb f/ft2) [32]. Erosion corrosion rates are also dependent on parameters
such as pH, dissolved oxygen content or other entrained gases, and temperature [64].
Increases in the pipe diameter will increase the thickness of the velocity boundary layer and the diffusion
boundary layer. The thicker velocity boundary layer leads to a decreased velocity gradient and reduced shear
stresses at the piping wall. For design purposes, a single velocity value marking the initiation of erosion
corrosion can provide misleading service life prediction. The maximum design velocities for 90/10 copper-
nickel piping decrease with piping size diameters below 6 inches and are shown in Table 1. The maximum
velocity can be as low as 34% of the generally accepted value (12 fps) for very small pipe diameters.
Removal of the corrosion product or oxide layer by excessive flow velocities leads to increased corrosion
rates of the metallic material. Corrosion rates are often dependent on fluid flow and the availability of
appropriate species required to drive electrochemical reactions. An increase in fluid flow may increase
corrosion rates by removing protective films or by increasing the diffusion or migration of deleterious species.
Conversely, increased flow may decrease corrosion rates by eliminating aggressive ion concentration or
enhancing passivation or inhibition by transporting the protective species to the fluid/metal interface.
Geometry

Mass transfer can have a significant effect on corrosion rates of metals and alloys depending on factors
such as bulk solution chemistry, temperature, and flow conditions, surface roughness, and geometry. Poulson
[65] showed that the mass transfer coefficient in two-phase annular flow in 180 elbows were 4-6 times that of
similar flow conditions in straight piping. Enayet et al. [66] measured the turbulent velocity profiles in a 90
elbow and found that the maximum corrosion rate and mass transfer occurred at the outer wall region.

8
TABLE 1 Maximum Design Velocities versus Pipe Diameter for
90/10 Copper-Nickel Alloy (UNS C70600)
Pipe Size Diameter (inches) Maximum Design Velocity
(feet/second)
4.2
4.8
1 5.4
1-1/4 6.2
1-1/2 6.6
2 7.4
2-1/2 8.2
3 9.1
3-1/2 9.8
4 10.3
5 11.5
6 and larger 12.0

Erosion corrosion of carbon steels in nuclear boilers has occurred downstream of flow disturbances such
as bends and orifices [67]. Factors that affect erosion in pipe bends include the degree of internal ellipticity and
asymmetry, sudden changes in cross-section, and reduction in cross-sectional area [67]. As the radii of
directional change are reduced, the erosion corrosion rate is expected to increase. Ledges, crevices, deposits,
other obstructions disturb laminar flow and result in turbulence at significant velocities [68]. Short distances less
than 10 diameters after directional changes or obstructions do not allow turbulence to dissipate [69]. Above a
critical Reynolds number, a given surface roughness increases mass transfer and the critical Re number
decreases with increasing roughness. Surface roughening tends to increase the mass transfer rate and
subsequent damage from erosion corrosion [70]. During testing, Poulson [67] observed that failures that
occurred on the bend intrados about 45 into an 180 copper pipe bend were associated with possible flow
separation while surface roughening only occurred at high Re numbers. Geometric factors related to different
configurations of piping components and fixtures are listed in [71].

Oxygen, pH, and Temperature

Boiler feedwater, boiler water, and condensate temperature influences the rates of both the oxidation and
reduction reactions. A plot of FAC versus temperature is a dome-shaped curve with a peak of about 300 F
(~150 C) for a given flow rate, steam quality and water chemistry [18]. Higher temperatures may also
increase corrosion rates by increasing the solubility of corrosion products. Increased oxygen and alkaline pH
stabilizes the oxide film on carbon steels and thus reduces FAC.
Lowering the pH in fresh water and seawater increases the rate of erosion corrosion at a given velocity
[64]. Low pH may prevent the repassivation or formation of passive films. Corrosion of copper alloys involves
anodic dissolution with an equal cathodic reaction, usually involving oxygen reduction. Deaeration of seawater
from 75 ppb to 5 ppb O2 substantially reduced erosion corrosion. Air and other gases present as bubbles in
flowing seawater augment the damage from erosion corrosion. The deleterious effect of the entrained bubbles is
decreased as the average diameter is decreased. Biocides such as chlorine or ozone can have an adverse effect
on the overall corrosion rates of stainless steels and nickel-based alloys. Temperature increases the reaction

9
kinetics of the cathodic and anodic reactions as well as oxygen diffusivity. However, increased temperatures
also decrease oxygen solubility in seawater and may increase biological activity. The surface shear stress
decreases for a given fluid velocity with increasing temperature [53].
As discussed earlier, many factors contribute to flow-assisted corrosion of metallic materials. Several case
histories that follow described some of the environmental and operational factors that contribute to flow-assisted
corrosion in naval systems.

CASE 1 - 90/10 COPPER-NICKEL PIPING

A 6x8 reducing spool (1150 gpm flow rate; specified minimum wall thickness 0.134) was a seam
welded 90/10 copper nickel pipe with a silver-brazed flange on both ends. Figure 2 shows the exterior of the
piping section. An area of through-wall penetration of the CuNi pipe occurred just past the outer 6 flange
sleeve at the inlet. Figure 3 shows the interior of the piping section in the area of the 6 flange. Heavy
waterside corrosion products around the inlet flange surface and associated material loss indicated severe
corrosion. A protruding lip of copper-nickel pipe relative to the flange alloy appears to indicate that corrosion
of the flange was caused by galvanic corrosion from its connection with an alloy 20 stainless steel seawater
pump. Through-wall penetration of the pipe was observed around the pipe circumference about 1 inch
downstream from the flange seat. General waterside corrosion and metal loss was heaviest in this area.
Shallow pit-like ripples of material loss, characteristic of impingement attack, were noted along the interior
surface in the area of pipe expansion from the 6-inch diameter to the 8-inch diameter piping. Several areas of
the waterside piping surface were lightly colored and displayed a lack of any scale due to erosion-corrosion.
Corrosion and general copper-nickel wall loss was also observed near the 8-inch flange at the outlet end of the
sample. Remaining copper-nickel wall thicknesses of 0.080-0.090 were measured near the outlet. Waterside
grooves and pits as deep as 0.025 were located on the outlet flange. Silicon detected in many of the deposit
samples suggests that sediment or debris participates in accentuating the erosive effects of seawater flow.
Although erosion-corrosion had occurred along the entire spool pipe length, galvanic corrosion of both flanges
as well as the more noble 90/10 copper-nickel piping in between also was observed in the overall failure. No
sulfides were detected in the remaining corrosion product scales.
Metallographic analysis of selected piping sections indicated typical microstructural features for 90/10
copper-nickel at both brazed and welded sites and in the general piping wall. Microhardness measurements did
not indicate any softening of the alloy which could lead to accelerated erosion corrosion effects. Chemical
analysis of the piping sample indicated that the pipe and flange conformed to UNS C70600, 90/10 copper-
nickel and M-bronze, respectively [57].
Piping welds exhibited virtually no loss of material from erosion corrosion as compared to adjacent 90/10
copper-nickel piping, as shown in Figure 4. Chemical analysis of piping welds revealed that they were
composed of copper-nickel alloy with a nickel content significantly higher than in the 10 wt. % nickel in the
piping. This higher nickel content in the weld deposit was also more resistant to galvanic corrosion since the
potential of the alloy composition is more electropositive than either 90/10 copper-nickel or bronze flange
materials.
In this case, protective copper hydroxy-chloride layers (Cu2(OH)3Cl) have attempted to form on the 90/10
copper-nickel reducing spool, but excessive seawater turbulence and flow rates exceeding the critical surface
shear stress removed the passive films from the copper-nickel alloy. Removal of these protective corrosion

10
product deposits increases galvanic corrosion due to contact with the more noble pump stainless steel alloy.
Both galvanic corrosion currents and turbulent flow acted synergistically to prevent reformation of adherent
corrosion product scales. This led to dissolution of the more susceptible, anodic 90/10 copper-nickel and
bronze alloys. The piping weld contained a higher nickel content than the piping alloy that both improved the
resistance of its corrosion product scales to erosion corrosion and decreased the driving force toward galvanic
corrosion because of its more electropositive potential compared to either bronze or 90/10 cooper nickel [57].

CASE HISTORY NO. 2 - 70/30 COPPER-NICKEL PIPING

A segment of a horizontal 70/30 copper nickel pipe had failed due to localized through-wall corrosion.
Water flowed through the pipe on a periodic basis; sometimes water sat in the pipe for several days before it
was circulated for disposal. The piping sample was a straight pipe section that had an elongated pit that was
between 0.4 and 0.8-in. wide and about 2-in. long; a hole about -in. diameter had perforated the pipe wall
near the center of this pit, Figure 5. Although the area in which the heavy corrosion occurred was elongated,
this was not strictly characteristic of erosion-corrosion. No significant corrosion was observed in other areas of
the pipe section. Following an initial visual inspection, the pipe sample was split longitudinally and 90 to the
plane in which the pitting had occurred. Figure 6 shows distinct differences in the corrosion product layers
between the failed side and the side opposite from the corrosion. The side on which the corrosion occurred
displays some spalling and disruption of the protective scale normally associated with copper-nickel piping in
seawater. The side opposite the failure had a green patina characteristic of the normal protective copper
hydroxy chloride scales and this side also did not display any significant corrosion. The different corrosion
product morphologies indicate that the through-wall penetration occurred at the bottom or 6 oclock position of
the pipe. The demarcation between the two scale morphologies may indicate the general level of water during
service. The bottom side around the failure showed little scaling within the corroded region and around the
perforation. Some residual sites of black scale remained on the surface around the corroded area; such sites
have been shown to be indicative of sulfide-induced corrosion.
Optical and scanning electron microscopic examination of several sites around the pipe did not indicate any
unusual metallurgical features or defects. Several locations were examined in the area where the severe wall
thinning occurred, including the lip of the through-wall pit and an area approximately -in. from the through-wall
pit. At these locations, the surface of the base material had a relatively uniform appearance with little or no
corrosion product scale observed. Minor intergranular corrosion of the waterside (interior) surface was
observed just beyond the large bottom pit. Although the interior pipe surface had experienced corrosion, those
corresponding surfaces were relatively smooth with minor intergranular corrosion. A porous scale about 100
m (0.004 in.) thick was present along the bottom pipe surface. The waterside surface beneath this black
deposit site (Figure 7) indicated that intergranular corrosion was more severe under the 100 m thick deposit.
These thick (100 m, 0.004 in.) black deposit scales were easily removed due to a lack of adherence to the
piping substrate and the observed degree of porosity could allow deleterious species to migrate through.
Removal of the scale would increase the local overall corrosion rate of the copper nickel pipe. Slight
intergranular corrosion (penetration less than 25 m (0.001 in.)) had occurred on the top pipe half where an
apparent protective scale about 40 m (0.002 in.) thick, (Figure 8) had formed. Analysis of the pipe chemistry
indicated that the piping was in conformance with specifications.

11
Elemental analysis of the general corrosion product scale by X-ray energy dispersive spectroscopy (EDS)
indicated the presence of chlorine (predominately chlorides) and other seawater-related constituents in the outer
film, while the sulfur content (as sulfides and/or sulfates) was below 1 wt % (Table 2). However, EDS analysis
of black deposit scales indicated an enrichment of phosphorus, sulfur, chlorine, calcium, and iron and a relative
decrease in nickel content in the waterside scale. It has been observed that phosphorus is a common nutrient
for many forms of bacteria including sulfate-reducing bacteria (SRB) that can produce sulfides in stagnant,
anaerobic environments [72], which can lead to accelerated corrosion of copper-nickel alloys.

Table 2 - EDS Semi-quantitative Analysis of Surface and Cross-Sectional Scales on


70/30 Copper-nickel Piping (weight percent)

Element General Black surface Black Cross-sect. Cross-sect.


pipe scale scale -site 1 Surface scale Bottom black Bottom ID scale
Site 2 scale - Site 3 @ pit Site 6
O --- --- --- 8.81 10.60
Na --- --- --- 1.47 1.24
Al 0.42 0.35 0.62 --- ---
Si 0.44 0.39 0.46 1.98 2.96
P 0.40 3.22 3.33 2.21 0.54
S 0.70 8.09 9.58 5.57 3.76
Cl 4.04 6.17 6.21 11.49 9.29
Ca 0.27 3.04 1.65 --- ---
Fe 1.78 5.86 4.81 2.60 1.97
Ni 22.54 5.22 7.30 5.76 11.87
Cu 69.42 67.65 66.04 60.10 55.28

Results of the analysis indicate that the corrosion in the pipe sample was primarily caused by poultice
corrosion enhanced by the presence of sulfides. The overall appearance of the corrosion product films indicates
that the large pit which ultimately perforated the pipe, was on the bottom of a horizontal piping run where debris
could readily accumulate. A ready source of organic material and the intermittent use of the system contributed
to the accumulation of deposits, leading to putrification and generation of sulfides. The sulfide-modified scales on
the copper-nickel pipe were less adherent; the periodic pipe flow tended to remove these less adherent scales
which subsequently accelerated dissolution of the copper-nickel pipe at these sites until corrosion product scales
could reform. Repeated cycles of scale formation, local sulfide modification of these scales by poultice
corrosion, and removal of these less adherent modified scales eventually led to failure at the bottom of the pipe.
The remaining surface scales along the bottom of the pipe, although different from the upper pipe scales, were
apparently sufficiently adherent to prevent more uniform corrosion along the lower pipe half. The relatively
smooth appearance of the large pit in the pipe sample suggests flow-assisted corrosion as a primary mechanism.
The overall service operation also suggests a chemically-assisted mechanism of accelerated corrosion by
generation of sulfides and modification of the corrosion product scales in the pitted area of this pipe.

12
CASE HISTORY NO. 3 UTILITY BOILER ROOF TUBE FAILURE

A nominally 0.221 thick roof tube from a utility boiler had failed reportedly from overheating due to
extremely heavy waterside deposits. This Combustion Engineering (CE) boiler, constructed in 1955, was
reportedly operating at 825 psig. One boiler is operated continuously for a one-year period in alternating years,
while maintenance is achieved on the other boiler unit. Present water treatment consists of a system of caustic,
phosphate, and polymer. Present feedwater and boiler components utilize only ferrous alloys.
The failed roof tube sample was removed and found to contain heavy waterside deposits on the hot side of
the tube. Subsequent examination of other roof tubes in the vicinity of this rupture revealed the lack of heavy
waterside deposits in corresponding areas of neighboring tubes. Previous analysis of boiler deposits from the
roof tube area indicated that the deposit contained a high (~ 80 wt. %) copper/copper oxide content, although
no utility boiler components were composed of copper-based alloys presently. Boiler water also was tested at
both intake and discharge and found to contain less than 1 ppm copper.
Figure 9 shows that the tube failure consisted of a one-inch longitudinal split at a bulge along the crown. A
side view of the respective waterside deposits (Figure 10) indicated that the hot side deposit was, in some
places along the tube sample, about one-inch thick. Deposits were mechanically removed in accordance with
Test Method A (mechanical removal) of ASTM 3483 (Standard Test Methods for Accumulated Deposition in
a Steam Generator Tube). The hot side deposit averaged 1000 g/sq.ft. The cold side deposit was
approximately 0.020 thick near the tube membrane and the deposit along the center of the cold side was less
than 0.005 thick. The cold side deposit averaged 62.8 g/sq. ft. Deposit analysis revealed high levels of copper
oxide and copper metal, amounting to 44 to 49 wt%. Magnetite, a normal oxidation product in boiler steels,
constitutes about 36 wt.%. Sodium and phosphate composition was about 1.8 %, while silica of about 10 wt.
% was detected.
Removal of the waterside deposits exposed the presence of random pitting. The nominal wall thickness for
the 3.0 I.D. tube is about 0.221. Pits on the cold side were 0.010 to 0.025 deep with the deeper pits
located along the tube membrane (which is not unusual for boiler tubing). Pitting along the hot side was 0.027
to 0.055 deep. General thinning of the tube wall at the failure reduced the general wall thickness to about
0.121-0.150. Pits around the failure region reduced the tube wall to a minimum of approximately 0.103.
The failed carbon steel tube conformed to ASME SA 192 or SA 210 chemical specifications. The
microstructure along the cold side (Figure 11) displayed a slightly banded configuration of ferrite and pearlite a
normal morphology of low carbon steel. The hot side microstructure away from the bulge displayed no thermal
degradation of the pearlite and ferrite matrix. The waterside deposit (0.024-0.028; 610-710 m thick) in the
vicinity of the failure exhibited several distinct layers of oxide (Figure 12). Some of the layers were quite dense
while others were very porous. One layer exhibited a dispersion of copper metal particles, while other layers
revealed needle-like, fibrous oxides. These thick, porous deposits of iron and copper oxides and copper metal
can have a relatively high insulating value that could provide opportunities where tube overheating could occur.
The tube microstructure near the failure (~ 0.5 away) displayed complete spheroidization of the carbides
from overheating. Voids developed at grain boundaries, particularly at three-grain junctions; the voids tend to
be oriented perpendicular to the tube surface (Figure 13). The formation of these voids is due to long-term
overheating and creep rupture. Some of the voids have grown and coalesced into larger voids and incipient

13
microcracks; some of the largest voids have formed surface oxides. There is slight elongation of the grains as a
result of tube swelling and expansion.
The porosity and the chemical composition of this heavy deposit in this horizontal roof tube tends to suggest
that the deposit slowly migrated to this site from other boiler sites over a long period of time. A lack of proper
coolant flow in this high-heat flux area generated further waterside deposition and caused local overheating
leading to creep rupture of the roof tube.

CONCLUSIONS
Flow-assisted corrosion can take many forms: (1) mass transport controlled corrosion, (2) phase transport
controlled corrosion, (3) erosion corrosion, and (4) cavitation. Flow-assisted corrosion is dependent on a
number of factors such as alloy composition, water chemistry, pH, biofouling, microbiologically influenced
corrosion, pitting and crevice corrosion, water pollution and contamination, alloy surface films, geometry and
surface roughness, galvanic interaction, fluid velocity and mode, oxygen content, heat transfer rate, and
temperature. A system designer often uses a single value to express the flow-assisted regime which a material
may be utilized - this may lead to misleading or misguided ramifications if the local environmental and operational
details are unknown.
Although flow-assisted corrosion is often caused by excessive velocities, the lack of sufficient fluid flow in
high-heat transfer areas also can lead to enhanced corrosion rates. Several case studies have been examined in
terms of the factors contributing to flow-assisted corrosion in naval related systems.

ACKNOWLEDGEMENTS

I gratefully acknowledge the efforts of Mrs. Rich Hays, Andrew Sheetz and the Metallurgical Laboratory
(Code 612) for the failure analysis of several naval components. The work was conducted in the Marine
Corrosion Branch (Code 613) under the direction of Mr. Robert J. Ferrara.

REFERENCES

1 - B. Poulson, Corrosion Science, v.23, 391 (1983).


2 - E. Heitz, Corrosion, v.47, 135 (1991); also CORROSION/90, paper 1, NACE International, Houston, TX
(1990).
3 - T. Sydberger, J. British Corrosion, v. 22, 83 (1987).
4 - G. Kuiper, Cavitation and Multiphase Flow Forum, J.W. Hoyt, O. Furuya, eds., v. 23, ASME, New York, 1
(1985).
5 - C.E. Brennen, Cavitation and Bubble Dynamics, Oxford Engineering Science Series 44, Oxford University
Press, New York, p.7 (1995).
6 - M. G. Fontana, Corrosion Engineering, 3rd ed., McGraw-Hill, New York, 104 (1986).
7 - R. T. Knapp, J.W. Daily, F.C. Hammitt, Cavitation, Engineering Societies Monograph, McGraw-Hill, New
York, 332 (1970).
8 - B. Poulson, Corrosion Science, v.35, 655 (1993).

14
9 - B. Chexal, J. Horowitz, R. Jones, B. Dooley, C. Wood, M. Bouchacourt, F. Remy, F, Nordmann, P. St. Paul,
Flow-Accelerated Corrosion in Power Plants, EPRI, Palo, Alto, CA and Electricite de France (1996).
10 - G. Butler, H.C.K. Ison, Proc. 1st International Congress on Metallic Corrosion, Butterworths, London
(1961).
11 - P.J. Boden, Corrosion Science, v.11, 353 (1971).
12 - D.B. Dewitt-Dick, M.C. Wangerin, CORROSION/96, paper 533, NACE International, Houston, TX (1996).
13 - J.H. Kim, Cavitation and Multiphase Flow Forum, O. Furuya, ed., v. 25, ASME, New York, 117 (1987).
14 - L. Tomlinson, C.B. Ashmore, J. British Corrosion, v.22, 45, (1987).
15 - R.D. Port, H.M. Herro, Nalco Guide to Boiler Failure Analysis, McGraw-Hill, New York, 58 (1991).
16 - Steam, 40th ed., S.C. Stultz, J.B. Kitto, eds., 42-13, Babcock & Wilcox, Barberton, OH, (1992).
17 - B. Chexal, J. Horowitz, R. Jones, B. Dooley, C. Wood, M. Bouchacourt, F. Remy, F, Nordmann, P. St. Paul,
Flow-Accelerated Corrosion in Power Plants, EPRI, Palo, Alto, CA and Electricite de France, 5-19 (1996).
18 - B. Chexal, J. Horowitz, R. Jones, B. Dooley, C. Wood, M. Bouchacourt, F. Remy, F, Nordmann, P. St. Paul,
Flow-Accelerated Corrosion in Power Plants, EPRI, Palo, Alto, CA and Electricite de France, 2-2 (1996).
19 - S.C. Dexter, C.H. Culberson, CORROSION/79, paper 227, NACE International, Houston, TX (1979).
20 - G.J. Danek, Naval Engineers J., v. 78, 763 (1966).
21 - T.P. Hoar, J. Applied Chemistry, v.11, 121 (April 1961).
22 - B. Phull, S. Pikul, R.M. Kain, Corrosion Testing in Natural Waters: Second Volume, ASTM STP 1300, (R.M.
Kain. W.T. Young, eds., 34, ASTM. Conshohocken, PA (1997).
23 - D.H. Pope, D.J. Duquette, A.H. Johannes, P.C. Wayner, Materials Performance, v.23 (4), 14 (April 1984).
24 - Biologically Induced Corrosion: Proceedings of the International Conference on Biologically Induced
Corrosion, S.C. Dexter, ed., NACE-8, NACE International, Houston, TX (1986).
25 - H.A. Videla, Manual of Biocorrosion, CRC Press/Lewis Publishers, Boca Raton, FL (1996).
26 - H.M Herro, R.D. Port, The Nalco Guide to Cooling Water System Failure Analysis, Chapter 6, McGraw-Hill,
New York, NY(1993).
27 - B.Little, F. Mansfeld, Proceedings of the H.H. Uhlig Memorial Symposium, PV 94-26, F. Mansfeld, A.
Asphahani, H. Bhni, R. Latanison ,eds., 42, The Electrochemical Society, Pennington, NJ (1995).
28 - J.P. LaFontaine, S.C. Dexter, Proceedings of the 1997 Tri-Service Conference on Corrosion, Session 6,
ONR, Wrightsville Beach, NC (November 1997).
29 - D.A. Shifler, D. Melton, H.P. Hack, CORROSION/98, paper 706, NACE International, Houston, TX (1998).
30 - K.D. Efird, T.S. Lee, Corrosion, v. 35, 79 (1979).
31 - J.P. Gudas, H.P. Hack, Corrosion, v. 35, 67 (1979).
32 - D.D. MacDonald, B.C. Syrett, S.S. Wing, Corrosion, v. 35, 367 (1979).
33 - B.C. Syrett, Corrosion Science, v. 21, 187 (1981).
34 - L.E. Eeiselstein, B.C. Syrett, S.S. Wing, R.D. Caligiuri, Corrosion Science, v.23, 223, (1983).
35 - J.P. Gudas, G.J. Danek, R.B. Niederberger, CORROSION/76, paper 76, NACE International, Houston, TX
(1976).

15
36 H.P. Hack, J. P. Gudas, DTNSRDC SME-79-85, CarderockDiv, Naval Surface Warfare Center,
(November, 1979).
37 - Localized Corrosion, in Corrosion-Metal/Environment Reactions, 2nd ed., v.1, L.L. Sheir, ed., Newnes-
Butterworths, London, 1:143 (1976).
38 - H.G. Fontana, Corrosion Engineering, 3rd ed., McGraw-Hill, New York, 51 (1986).
39 - R.T. Jones, Carbon Steels and Low Alloy Steels, in Process Industry Corrosion - Theory and Practice, B.J.
Moniz, W.I. Polluck, eds., NACE International, Houston, TX, 381 (1986).
40 - C.H. Hare, J. of Protective Coatings and Linings,v. 15 (1) 67 (1998).
41 - M.G. Fontana and N.D. Greene, Corrosion Engineering, 2nd ed, McGraw-Hill, New York, NY, 39 (1978).
42 - P.T Gilbert, 3rd ed., L.L Sheir, R.A. Jarman, G.T. Burstein, eds., Butterworth-Heinemann, Oxford, 4:53
(1994).
43- H.P. Hack, H. Shih, H.W. Pickering, Surfaces, Inhibition, and Passivation, E. McCafferty and J. Brodd, eds.,
Proceedings Volume 86-7, The Electrochemical Society, Penington, NJ, 355 (1986).
44 - H.P. Hack, DTNSRDC SME-87-22, CarderockDiv, Naval Surface Warfare Center, (November 1987).
45 - C. Kato, J.E. Castle, B.G. Ateya, H.W. Pickering, J. Electrochemical Society, v. 127, 1897 (1980).
46 - C. Pearson, British Corrosion Journal, v. 7, 61 (1972).
47 - R.F. North, M.J. Pryor, Corrosion Science, v. 10, 297 (1970).
48 - G.L. Bailey, J. Institute of Metals, v. 79, 243 (1951).
49 - W.C. Stewart, F.L. LaQue, Corrosion, V. 8, No. 8, 259 (1952).
50 K.D. Efird, Corrosion, v. 33, no. 10, 347 (1977).
51 - L.J. P. Drolenga, F.P. Ijsseling, B.H. Kolster, Werkstoffe und Korrosion, v. 34, 167 (1983).
52 J.P. Frick, L.R. Scharfstein, T.M. Parrill, G. Haaland, J. of Metals, v. 37, No. 3, 28 (1985).
53 K.D. Efird, Corrosion, v. 33, no. 1, 3 (1977).
54 - J.W. Oldfield, Electrochemical Theory of Galvanic Corrosion, Galvanic Corrosion, H. P. Hack, ed.,
ASTM, West Conshohocken, PA, p. 5, 1988.
55 - R. Francis, British Corrosion Journal, Vol. 29, p. 53, 1994.
56- H.P. Hack, Galvanic Corrosion of Piping and Fitting Alloys in Sulfide-Modified Seawater,
DTNSRDC/SME-79-31, (May, 1979).
57 - F.L. LaQue, Marine Corrosion, Wiley-Interscience, New York, 179 (1975).
58 - G.A. Gehring, C. K. Kuester, J.R. Maurer, CORROSION/80, paper 32, NACE International, Houston, TX
(1980).
59 - G.A. Gehring, J.R. Maurer, CORROSION/81, paper 202, NACE International, Houston, TX (1981).
60 - D.J. Astley, Corrosion Science, v.23, 801 (1983).
61 - J.R. Scully, H.P. Hack, Prediction of Tube-Tubesheet Galvanic Corrosion Using Finite Element and Wagner
Number Analyses, Galvanic Corrosion, ASTM STP 978, H.P. Hack, ed., ASTM, West Conshohocken, PA,
136 (1988).
62 - K.D. Efird, Corrosion, v. 33, 3 (1977).

16
63 - J.M Popplewell, CORROSION/78, paper 21, NACE International, Houston, TX (1978).
64 - B.C. Syrett, Corrosion, v.32, 242 (1976).
65 B. Poulson, Chemical Engineering Science, v. 46, 1069 (1991).
66 - M.M. Enayet, M.M. Gibson, M. Yianneskis, International J. Heat & Fluid Flow, v. 3 113 (1982).
67 - B. Poulson, R. Robinson, International J. Heat & Fluid Flow, v. 31, 1289 (1988).
68 M.G. Fontana, N.D. Greene, Corrosion Engineering, 2nd ed, McGraw-Hill, New York, NY, 78 (1978).
69 N. S. Hirota, Erosion-Corrosion in Wet Steam Flow, Metals Handbook, Corrosion, 9th ed, v. 13, 967, ASM
International, Materials Park, OH (1987).
70 - B. Poulson, Corrosion Science, v. 30 743 (1990).
71 - B. Chexal, J. Horowitz, R. Jones, B. Dooley, C. Wood, M. Bouchacourt, F. Remy, F, Nordmann, P. St. Paul,
Flow-Accelerated Corrosion in Power Plants, EPRI, Palo, Alto, CA and Electricite de France, 4-11 (1996).
72 - A.A. Stein, MIC Treatment and Prevention, in A Practical Manual on Microbiologically Influenced Corrosion,
G. Kobrin, Ed. NACE International, Houston, TX, 101 (1993).

Figure 1 Macrofouling of titanium pipe in strained, flowing (6 fps, 1.8 m/s) seawater after 6
months.

17
Figure 2 8 x 6 Reducing Spool Inlet at 6-inch end. Arrow points to region of failure.

Figure 3 Severe waterside corrosion of inlet flange and first inch of copper-nickel pipe
observed. Less severe corrosion affected the remainder of copper-nickel pipe and outlet
flange.

18
Figure 4 Cross-section of piping weld deposit (dark area) contains 16-19 wt. % nickel and
exhibits very little loss from either erosion corrosion or galvanic corrosion while the 90/10
copper-nickel pipe has lost 2000-2500 m (~ 0.080-0.100) of its original 0.134 wall
thickness.(light areas). 10X.

Figure 5 70/30 copper-nickel piping sample as received.

19
Figure 6 Piping sample after sectioning shows distinct differences in corrosion product films of
the bottom pipe half (left side) and the upper pipe half (right side).

Figure 7 Pipe surface beneath black deposit scale.

20
Figure 8 Pipe surface beneath normal protective scale along upper pipe half.

Figure 9 Roof tube as received. Narrow one-inch long failure is along the hot side crown.

21
Figure 10 Side view of tube sample reveals that hot side deposits are as much as 1.0 thick.
Relatively little deposition was present on the cold side.

Figure 11 Microstructure along cold side shows the lamellar carbides in the pearlite
colonies. Spherical phases are small alloy constituents such as oxide and sulfide
inclusions. Magnification - 1050X.

22
Figure 12 Hot side deposit in the vicinity of the failure displays several layers and waterside
oxidation/corrosion. Some layers are dense while others are very porous. One layer displayed
needle-like, fibrous oxides. Another layer contains a number of particles of copper metal.
Magnification - 210X.

Figure 13 Carbides are fully spheroidized from thermal degradation near failure. Voids (dark
sites) have formed along the grain boundaries that are perpendicular to the direction of applied
stress. Magnification - 1050X.

23

S-ar putea să vă placă și