Tesis Corrosion, Codigos Phreeqc

Preface
This report is the result of my graduation research for the MSc track Water Management at the
Delft University of Technology. Looking back, I can affirm that studying at the TU Delft has
taught me a lot, opened many doors and led to invaluable experiences. Finishing my MSc studies
would not have been possible without the support that I received.
Firstly, I would like to thank Luuk Rietveld for encouraging me to pursue my studies in the section
of Sanitary Engineering. Secondly, regarding my graduation work, I would like to thank several
people that have been part of the process. Henri Spanjers, you introduced me in the world of
industry water and remained a patient supervisor despite the time I took to finish the work. David,
thank you for enabling the partnership with Evides Industriewater. It was absolutely interesting to
assist you with the pilot cooling tower in Terneuzen. Both David and Ingrid, thank you for your
support and for continuing to keep me updated about the second pilot. Furthermore, Jules van Lier
and Bas Heijman, thank you for your lectures and for being part of my graduation committee.
Additionally, I'm very grateful to Arjan Mol, you welcomed me in the Corrosion Technology and
Electrochemistry group. Mats and Joost, thank you for guiding me in the corrosion labs, helping
out with the potentiostats and making the lab life more sociable together with the other Rusty
People. The Waterlab and the faculty of Mechanical, Maritime and Materials Engineering, with
all the workshops and lab facilities, have been a true play ground to me. This can be credited to
the support of the (lab) technicians, Mohammed, Armand, Agnieszka, Sander and Wim.
At last, I am thankful for my family and friends, who provided the distraction and the help that I
was seeking for. Charlotte, thank you for showing up when needed most and for encouraging me
to take the programming one level up. Maurits, I owe you great gratitude, I feel privileged with
your genuine support and your eagerness to understand my work.
Soli Deo gloria.
Wyona Boers
Delft, August 2018
iii
Abstract
Water as a coolant for industrial products or processes is widely used. Efficient operation of Open
Recirculating Cooling Water Systems (ORCWS) relies on controlled cooling water conditioning
programs that prevent corrosion, fouling, scaling and microbial growth. Water intensive industries
feel greater pressure to decrease their water demand as water resources become increasingly
stressed and discharge restrictions become more stringent. Pre-treating the cooling water before it
enters the recirculation stream provides the opportunity to reduce the water demand and to seek
for alternative chemicals, preferably less complex and with a lower stress on the environment. Soft
alkaline cooling water (SACW) might be a good alternative for the conventional water and
chemicals intensive ORCWS operation. However, this type of water requires complete
reevaluation of the conditioning programs. Therefore, this research focused on the corrosion
processes in SACW and a beginning was made in evaluating non-toxic and phosphorus-free
corrosion inhibitors.
The research comprised the production of synthetic cooling water (SCW) and the evaluation of
corrosion of mild steel in this type of water, with and without inhibitors. The selected corrosion
inhibitors were NovoTraqua® without phosphorus (NT0) and sodium silicate (Si200). PHREEQC
simulation were used to validate the SCW composition. A desk study was carried out to evaluate
the applicability of corrosion indices. The average corrosion rate and inhibition efficiency were
determined by mass loss tests of mild steel specimens with an immersion time of up to 4 weeks.
The development of passivity was followed by open circuit potential (OCP) measurements. The
pitting susceptibility was tested using cyclic potentiodynamic anodic polarization (CPP)
measurements. 3D digital microscopy, SEM/EDS and XRD were used to evaluate the surface
morphology and the composition of the top layer.
The calculations for the SCW resembled PHREEQC simulations. A natural pH of 9.1 was reached
based on the carbonate equilibrium. Small corrections in the NaHCO3 and Na2CO3 dosage were
applied to increase the pH to the desired pH 9.4. The existing corrosion indices are not suitable to
be applied to SCW. The immersion of mild steel with a roughness of 30μm in SCW resulted in
uniform corrosion. The addition of NT0 or Si200 to the SCW resulted in the inhibition of uniform
corrosion but no complete protection against localized corrosion was realized. Si200 was the best
performing inhibitor. The immersion of mild steel with a 7μm roughness in SCW resulted in
natural passivation by the formation of a Fe2O3 oxide layer. The SCW did not have the ability to
immediately repassivate the specimens after localized breakdown of the protecting passivating
layer. The addition of NT0 or Si200 reduced the aggressiveness of the environments at the pits but
did not accomplish complete repassivation.
Lab experiments with SCW can provide initial insights in the behavior of mild steel and can
contribute to the decision making in more cost intensive research on alternative operation of
ORCWS. The experiments should be extended by testing the behavior and inhibition of corrosion
at different temperatures, inhibitor concentrations, phosphonate concentrations and different
v
Abstract
cycles of concentration. Better understanding of the protection mechanisms would be obtained

when more advanced techniques like Fourier Transform Infrared Spectroscopy (FTIR) and
Electrochemical Impedance Spectroscopy (EIS) are used. Although cooling water corrosion
inhibition by means of silica is a mature research topic, it is recommended to proceed with research
on silica as a corrosion inhibitor in the specific conditions of SACW.
vi
Contents
Contents
Preface .......................................................................................................................................... iii
Abstract .......................................................................................................................................... v
Contents ....................................................................................................................................... vii
List of Figures ............................................................................................................................... ix
List of Tables ................................................................................................................................ xi
1 Introduction .......................................................................................................................... 13
1.1 Background and Problem description ...................................................................... 13
1.2 Objective and Research Questions ........................................................................... 15
1.3 Research approach and Report Outline .................................................................... 16
2 Theoretical background ........................................................................................................ 17
2.1 Open recirculating cooling systems ......................................................................... 17
2.2 Corrosion in aqueous environments ......................................................................... 19
2.3 Corrosion measurements .......................................................................................... 25
3 Synthetic Cooling Water ...................................................................................................... 33
3.1 Materials and Methods ............................................................................................. 33
3.2 Results and Discussion............................................................................................. 37
4 Mass loss tests ...................................................................................................................... 41
5 Voltamettric tests ................................................................................................................. 51
6 General discussion and Conclusions .................................................................................... 65
7 Recommendations ................................................................................................................ 69
8 Bibliography......................................................................................................................... 71
9 Abbreviations ....................................................................................................................... 77
vii
Contents
Appendix A – HumVi production and composition..................................................................... 79

Appendix B – PHREEQC input file ............................................................................................. 81
Appendix C – 4 weeks immersion test setup................................................................................ 83
Appendix D – Specimen composition and surface roughness ..................................................... 85
Appendix E – Corrosion cell setup............................................................................................... 87
Appendix F – CPP test results ...................................................................................................... 89
Appendix G – Surface visualization after CPP tests .................................................................... 93
Appendix H – XRD results .......................................................................................................... 95
viii
List of Figures
Figure 1: Tetrahedron of undesired interacting processes in ORCWS. ....................................... 13
Figure 2: Schematization of an Open Recirculating Cooling Water System. .............................. 17
Figure 3: Makeup rate and blowdown as a function of COC. ..................................................... 18
Figure 4: Schematic corrosion cell of iron in an aqueous environment. ...................................... 19
Figure 5 Simplified theoretical potential-pH diagrams for iron in water..................................... 20
Figure 6: Variation in the cross-sectional shape of pits. .............................................................. 22
Figure 7: Classification of inhibitors ........................................................................................... 24
Figure 8: Hypothetical cathodic and anodic polarization diagram. ............................................. 29
Figure 9: Hypothetical cathodic and anodic polarization plots for a passive anode .................... 29
Figure 10: Hypothetical CPP diagrams ........................................................................................ 30
Figure 11: PHREEQC input solution and output characteristics ................................................. 39
Figure 12: Impact of an increasing COC of SCW on the Saturation Index (SI) and the pH ....... 39
Figure 13: Schematic representation of the 4 weeks immersion test setup of a single jug .......... 43
Figure 14: Photo of the 4 weeks immersion test setup................................................................. 44
Figure 15: 2L Measuring jug with custom made lid. ................................................................... 44
Figure 16: Custom made lid ......................................................................................................... 44
Figure 17: Inhibitor concentration development .......................................................................... 45
Figure 18: Corrosion rates for Blanc, CAO, NT100 and Si200 at a weekly interval................... 48
Figure 19: Results of surface analysis after immersion ............................................................... 50
Figure 20: Corrosion cell set-up, exploded view with water. ...................................................... 52
Figure 21: Corrosion cell set-up, assembled view without water. ............................................... 53
Figure 22: OCP Development of mild steel immersed in SCW. ................................................. 55
Figure 23: Final OCP values after 6 days of immersion in different solutions. ........................... 56
Figure 24: Three types of CPP curves ......................................................................................... 57
Figure 25: Overview of representative CPP curves. .................................................................... 59
Figure 26: Overview of the breakdown points of the CPP tests .................................................. 60
Figure 27: Digital microscopy image at 200x of crevice corrosion at the polyester interface..... 61
Figure 28: Digital microscopy images of a pit after a CPP test ................................................... 62
Figure 29: The maximum depth to average width ratio of pits after CPP tests. .......................... 63
Figure 30: SEM/EDS results ........................................................................................................ 64
Figure 31: Schematization of the 4 weeks immersion tests setup. ............................................... 83
Figure 32: Surface roughness of Metal Samples Co coupon vs. self-made coupon. ................... 85
Figure 33: Line drawing of self-made corrosion cell ................................................................... 87
Figure 34: Overview of the CCP tests per solution type .............................................................. 89
Figure 35: Overview of the CCP tests ......................................................................................... 90
Figure 36: Overview of the surfaces after CCP tests per solution type ...................................... 93
Figure 37: XRD measurements of mild steel coupons after immersion ...................................... 95
ix
List of Tables
Table 1: Interpretation guidelines for the LSkI ............................................................................ 26
Table 2: Interpretation guidelines for the RI ................................................................................ 26
Table 3: Main water composition of the Dutch surface waters.................................................... 34
Table 4: Chemicals used to create the stock solutions to produce SCW. .................................... 35
Table 5: SCW composition with a COC of 10............................................................................. 38
Table 6: Dosages for SCW with a COC of 10. ............................................................................ 38
Table 7: Rating table for pit density............................................................................................. 46
Table 8: XRD results ................................................................................................................... 63
Table 9: Composition of HumVi ................................................................................................. 79
Table 10: Chemical composition of the specimens...................................................................... 85
Table 11: Overview of breakpoint or vertex indices, current densities and potentials. ............... 91
xi
1 Introduction
1.1 Background and Problem description

Water as a coolant for industrial products or processes is widely used considering its availability
in most industrialized areas, its nontoxicity, its relatively high heat capacity, and its thermal
conductivity. The heat from the hot process fluids is transferred to the cooling water in a heat
exchanger and carried away by the cooling water. The performance of a cooling water system
(CWS) is critical for the production process. Failure of the cooling system can result in decreased
heat transfer, economic losses due to unscheduled shut downs, and failure of the production
equipment.
The concentration of the dissolved and suspended solids, and the contamination of the cooling
water through interaction with the atmosphere, cause highly undesired processes for Open
Recirculating Cooling Water Systems (ORCWS). Corrosion of metal, formation of scale, fouling
of the system, and bacterial growth are all undesired processes, that also influence and intensify
each other. This mutual interaction is schematized in Figure 1, and an example would be that when
insoluble corrosion or scale products aggregate and settle, it results in fouling and occluded areas,
which again results in growth sites for microbiology and under deposit corrosion. ORCWS
efficient operation depends on controlled cooling water conditioning programs to counteract these
processes that occur simultaneously. Conventional conditioning programs rely on a tailored
mixture of corrosion, scaling and fouling inhibitors. Toxic chemicals, such as chromate, were
commonly used in corrosion inhibitors for cooling water conditioning programs until prohibited
by legislation. Current prevailing alternatives for those toxic inhibitors are phosphorus containing
inhibitors. [1,2]
Figure 1: Tetrahedron of undesired interacting processes in ORCWS.
13
Introduction
A robust and reliable operation of the ORCWS is crucial, and opportunities to reduce the
operational costs and the environmental impact are explored. Water intensive industries feel
greater pressure to decrease their water demand as water resources become increasingly stressed
and discharge restrictions become more stringent. The cooling towers and infrastructure may often
not have reached their end of life, therefore redesigning the cooling water conditioning can provide
possible solutions. Conditioning the cooling water before it enters the recirculation stream
provides the opportunity to reduce the water demand and to seek for alternative inhibitors,
preferably less complex and with a lower stress on the environment.
One of the options to condition the water before it enters the ORCWS is to apply complete
softening. Softening is about the removal of bivalent cations, calcium and magnesium ions in
particular. By applying softening as a pretreatment step, the risk of scaling of the ORCWS is
reduced and the opportunity to operate at higher concentration ratios is established. Besides, the
recirculating cooling water in an ORCWS develops a natural high pH due to the gas exchange
with atmosphere. Common practice, to reduce the risk of scaling, is to correct the pH to a value in
the range of pH 7 to pH 8.5 with acids or to apply an All-Organic inhibition program that contains
phosphonates and which is effective at an elevated pH. So without the presence of scale forming
ions, it would be possible to operate the ORCWS at a higher concentration ratio and at a naturally
established pH, resulting in a decrease of the water demand, of the loss of inhibitors, and of the
need for pH correcting acids or the need for complex phosphorus containing inhibition programs.
Although operating a ORCWS with soft water mitigates the risk of scaling, other undesired
processes, like corrosion, will still occur and require inhibition. Prevailing corrosion inhibitors for
soft alkaline cooling water (SACW) are zinc, molybdate, phosphonates and organic phosphates
[1]. However, non-disrupting inhibitors are characterized by non-toxicity and macro nutrient
limitation, which is not applicable to the corrosion inhibitors just mentioned. Therefore,
sustainable operation of ORCWS with SACW requires a reevaluation of the suitable corrosion
inhibitors.
Limiting phosphorus has positive consequences for the water quality. Firstly, high phosphorus
concentrations in the reject water disturb the environment by causing eutrophication. The
maximum phosphorus concentration is restricted by legislation and is expected to become even
lower by law in the future. Consequently, the reject water requires advanced treatment before
being discharged to surface waters. Secondly, phosphorus is also a macronutrient and vital for the
undesired microbial growth in the ORCWS. The microbial growth is not only regulated by organic
carbon but also by the phosphorus and nitrogen concentration [3]. New technologies, like
BioPhree® [4], are being developed to remove phosphorus efficiently from pre-treated water. The
combination of a high pH and limited phosphorus content may be an alternative for the usage of
biocides to reduce the microbial activity in ORCWS.
According to a desk study executed by KWR Watercycle Research Institute (KWR) [5], it is
expected to be cost and environmentally advantageous to apply complete softening before the
water enters the recirculation stream of the ORCWS. Operating ORCWS with soft water is not a
14
Introduction
new phenomenon to the industry but corrosion remains a complex process that is not without
ambiguity and requires research [6]. Water engineers are often confident in assessing the scaling
potential, for which indices and rules of thumb have been developed. Attempts have been made
to come up with corrosion indexes but those are limited in their usage range [5,7–10].
Corrosion research can focus on a great variety of metals and alloys. A widespread and economical
construction material for cooling water systems is mild or low-carbon steel. Corrosion research
on mild steel for ORCWS has been performed in soft water, albeit at (near) neutral pH, and in
more alkaline water, but which contained calcium and magnesium ions [11–17]. Research
focusing on corrosion in concrete or in boiler systems does utilize soft alkaline solutions.
Nevertheless, this research generally excludes the presence of oxygen [18]. A respectable amount
of corrosion research on construction metals in aqueous environments is related to the oil and gas
industry and focusses on very saline conditions, on acidic conditions, and on carbon dioxide
corrosion [19,20]. Nor do those conditions prevail in ORCWS. So more knowledge is to be gained
on the corrosion behavior of mild steel in SACW to accomplish more sustainable operation of
ORCWS, as was recommended by KWR [5].
1.2 Objective and Research Questions

This thesis focusses on the corrosion processes in SACW and a beginning is made in evaluating
non-toxic and phosphorus-free corrosion inhibitors. The main goal of this thesis is:
“To gain understanding of the corrosion behavior of mild steel in soft alkaline cooling water, to
test the effect of different corrosion inhibitors on the corrosion behavior, and to explore
phosphorus-free corrosion inhibitors for this type of cooling water”.
This goal is reached in such a way that the results contribute to the decision making and
understanding of other (pilot scale) research projects that support the operation of ORCWS
without the addition of toxic or phosphorus containing inhibitors.
Exploring the different types of corrosion assessment techniques and selecting the appropriate
tests for this thesis is key. This results in the first research question:
A. How can the corrosion behavior of mild steel in soft alkaline cooling water be evaluated?
The selected tests and test conditions are first of all used to discuss the effect of only SACW on
the corrosion of mild steel before the effect of different inhibitors is tested. Therefore, the
following two questions are posed:
B. What is the effect of soft alkaline cooling water on the corrosion behavior of mild steel?
C. How do different non-toxic corrosion inhibitors affect the corrosion behavior of mild steel
in soft alkaline cooling water?
15
Introduction
It is expected from the answers to the previous questions that a corrosion inhibitor is required to
establish a sufficient protective layer on the surface of mild steel. The next step is to evaluate if
the corrosion inhibitor can be protective without the presence of additional phosphorus.
D. How do different concentrations of phosphorus in the corrosion inhibitor affect the

corrosion behavior of mild steel in soft alkaline cooling water?
1.3 Research approach and Report Outline

The report provides first of all, in Chapter 2, some theoretical background on CWS, on corrosion
in aqueous environments, and on different techniques and methods that can be applied to evaluate
corrosion. This is based on literature review. Chapter 3, 4, and 5 each cover a different
experimental phase. Therefore, each of these three chapters is split up into ‘materials and methods’
and ‘results and discussion’. Chapter 3 describes what kind of cooling water is used for the
experiments and what the characteristics are. This water is used for the actual corrosion
experiments that are described in Chapter 4 and 5. Chapter 4 evaluates corrosion in different
solutions based on mass loss experiments, while Chapter 5 evaluates corrosion in different
solutions based on potential and current density measurements. Finally, the research questions
presented in the previous section are answered in Chapter 6 and the recommendations that arose
from the research are provided in Chapter 7.
The relevance of this research is directed at open recirculating cooling water systems, which will
be for remainder of thesis referred to as cooling water systems or CWS.
16
2 Theoretical background
2.1 Open recirculating cooling systems

In CWS, the heat of the cooling water is released through evaporation. As a result, the dissolved
and suspended solids in the cooling water become more concentrated. The water that leaves the
cooling tower by evaporation and other losses is replenished with unconcentrated water, called
makeup water, to avoid further concentration increase. The solids concentration in the cooling
water is regulated by controlled discharge of this water as blowdown. The different water flows
and losses are illustrated in Figure 2. The ratio of an ion concentration in the blowdown (Cbd) to
the concentration of the same ion in the makeup water (Cmu) is defined as cycles of concentration
(COC) or the concentration ratio, as in Equation (2.1).
Cbd
COC  (2.1)
Cmu
Figure 2: Schematization of an Open Recirculating Cooling Water System.
17
Theoretical background
The blowdown rate (BD) and the makeup rate (MU) depend on the evaporation rate (E) and the
desired COC, as is given in Equation (2.2) and (2.3) and illustrated in Figure 3. Operating at a
higher COC reduces the blowdown rate. The makeup rate is asymptotically limited by the
evaporation rate, because the evaporation rate is dependent on the conditions of the system and
the atmosphere.
E
BD  (2.2)
(COC  1)
E  COC
MU  (2.3)
(COC  1)
Figure 3: Makeup rate and blowdown as a function of COC.
The concentration of ions and inhibitors in a CWS is carefully monitored and regulated. The
holding time index (HTI) is defined as the time required for the concentration of an ion to dilute
to half of its original concentration within the systems volume (V). It is calculated with Equation
(2.4).
ln(2)  V
HTI  (2.4)
BD
Increasing the COC, when initially operating the CWS at a COC up to five, will reduce the water
demand. Increasing the COC above six has no effect on the water demand but it does still reduce
the volume of the waste stream and the amount of chemicals that is lost.
18
2.2 Corrosion in aqueous environments
2.2.1 Thermodynamics and kinetics of iron corrosion

The deterioration of metal in an aqueous environment in case of corrosion is an electrochemical
process where the metal returns to its natural oxide state. Corrosion occurs if an electrochemical
corrosion cell is established. This cell comprises an anode, a cathode, a ionic current path, and an
electronic path, as is shown schematically in Figure 4. The anode generates electrons and
positively charged metal ions in an oxidation reaction. The cathode consumes electrons in a
reduction reaction.
The anodic oxidation reaction of a metal in general:
M  M n   ne (2.5)
The cathodic reduction reaction in natural waters of neutral or alkaline pH, is dominated by the
oxygen reduction reaction:
O2  2H 2O  4e  4OH  E0 = + 0.40 V SHE (2.6)
The ionic current path is established in a conducting electrolyte. The electronic path transfers the
electrons from the anode to the cathode, establishing an electric current. The driving force of the
electrochemical corrosion reaction is the electrical-potential difference between the anode and the
cathode, and causes current to flow between the anode and the cathode.
Figure 4: Schematic corrosion cell of iron in an aqueous environment.
The electrochemical processes can be displayed in a thermodynamic framework based on

equilibrium equations. Based on these thermodynamic equilibria, the electrode potential can be
plotted in a graph against a chosen factor, pH or temperature for example. The graph will show
the conditions of thermodynamic stability based on the lowest free-energy state for the reactions
considered. Figure 5 is a simplified potential-pH equilibrium diagram for iron-water from
Pourbaix [21].
19
Figure 5 Simplified theoretical potential-pH diagrams for iron in water, with the experimental equilibrium potentials
in the absence and in the presence of oxygen (based on diagrams in [21]).
Three type of domains or particular types of stability can be distinguished, namely corrosion
(stability of a dissolved species), immunity (stability of the metal), and passivity (stability of solid
oxides or salts) [22]. The dashed lines ‘a’ and ‘b’ indicate the stability domain of water, which is
defined by the electrochemical reactions of hydrogen evolution and oxygen reduction according
to Equation (2.7) and (2.8).
2H   2e  H 2 E0 = 0.0 V SHE (2.7)
O2  4H   4e  2H 2O E0 = + 1.23 V SHE (2.8)
Series of experiments have been performed to examine the behavior of iron in various aqueous
solutions and under various conditions of pH and electrode potential to determine domains of
corrosion, immunity and passivation, and to compare this with the theoretically established
diagrams [21]. Figure 5 represents the results from experiments by Pourbaix [21] of iron in
solutions free from oxygen and in solutions saturated with oxygen at atmospheric pressure. The
presence of oxygen increased the potential of iron. The potential of iron is even raised to the
passivation domain at pHs above 8. Evans [23,24] had already proven that iron, in a solution that
contains oxygen, is passivated by an oxide film and that differential aeration increases the
corrosion rate in the non-aerated regions. Mayne and Pryor [25,26] reported that the electrode
potential of iron increases in the presence of oxygen in the solution. In an aerated solution, a thin
film of Fe2O3 is formed, which institutes the development of positive potentials. It is believed that
this film eventually becomes impervious to oxygen and to ions, establishing protection against
corrosion.
20
Thermodynamic calculations or potential-pH diagrams are useful as a first study of the expected
stability domain, but do not provide information on the reactions rates or the transformation
process from iron to an iron oxide. The two common oxidation states for iron are Fe(II) and Fe(III),
ferrous and ferric iron, respectively. Ferric iron is the stable oxidation state in aqueous solutions
in equilibrium with the atmosphere.
Before stable oxide products are formed, the transformation of the iron in the steel surface
undertakes different electrochemical reactions. The Fe2+ ion is the initial product created by the
anodic dissolution reaction of steel:
Fe  Fe2  2e E0 = -0.44 V SHE (2.9)
which can either form simple salts with most anions or it forms the first hydroxide, Fe(OH) 2
(ferrous hydroxide). The reaction in Equation (2.9) is complemented by the cathodic reduction
reaction of oxygen given in Equation (2.8), resulting overall reaction:
2 Fe  O2  2 H 2O  2 Fe(OH ) 2 (2.10)
Fe(OH)2 has a low solubility product, as is given in Equation (2.11). Therefore, Fe(OH)2 is formed
close to the metal surface and deposited on the surface. Ferrous hydroxide is however unstable in
a solution that is in equilibrium with the atmosphere and is therefore further oxidized to form
Fe2O3  nH2O (hydrous iron(III) oxide), FeO(OH) (iron(III) oxide hydroxide), Fe2O3 (anhydrous
iron(III)oxide) or Fe3O4 (magnetite). Fe3O4 is stable in oxygen limited environments. The reaction
to form Fe2O3  nH2O is given in Equation (2.12).
K s  (aFe2  )( aOH  ) 2  1.64  10 14 at 18°C (2.11)
2Fe II (OH )2  2OH   Fe III 2O3  H 2O  2H 2O  2e (2.12)
FeO(OH), Fe2O3 and Fe3O4 can form coherent adherent layers on the metal that passivate the metal
by inhibiting the transport of reacting species between the metal and its environment, called a
passivating layer or passive film. Cohen [27] proposed several models of passive film on iron and
Strehblow [28,29] confirmed that for iron an inner Fe3O4 film is followed by Fe2O3 with the
possibility of FeO(OH) on top in alkaline solutions.
For the formation of a protective passivating layer it is key that the ferrous ions precipitate
immediately and rapidly at the surface and get further oxidized to protective ferric compounds.
This requires a sufficient supply of oxygen. This supply of oxygen to the metal surface layer is
slow unless the velocity of water relative to the metal surface is sufficiently high. The corrosion
of steel in air-saturated water is under diffusion control of dissolved oxygen because of the limiting
concentration and diffusion coefficient of oxygen, and because of the rapid corrosion reaction
when oxygen reaches the steel surface. The reaction rate of the oxygen reduction reaction is
strongly pH dependent and is highest in an alkaline environment. The unavailability of dissolved
oxygen due to limiting diffusion can result in unprotective Fe2+ products.
21
2.2.2 Localized corrosion

Two types of localized corrosion are pitting corrosion and crevice corrosion. Pitting is a highly
localized form of attack that often results in small anodic sites where holes are created that can
penetrate deep into the metal. Failure of the passive film or the formation of concentration cells
under a deposit can result in pitting. Pitting can be induced by halide ions, most commonly by the
Cl- ion. Its unpredictable behavior can cause sudden failure of the metal. Crevice corrosion occurs
in crevices where the solution is often stagnant. Figure 6 provides an overview of the possible pit
shapes.
Figure 6: Variation in the cross-sectional shape of pits. (a) Narrow and deep. (b) Elliptical. (c) Wide and shallow.
(d) Subsurface. (e) Undercutting. (f) Shapes determined by microstructural orientation. [30]
2.2.3 Water parameters affecting corrosion

Although corrosion of mild steel in aerated water proceeds by the action of dissolved oxygen, the
rate of the reaction is affected by multiple parameters. The water quality parameters that especially
affect the corrosivity of water are pH, DO, hardness, alkalinity, TDS, natural organic matter
concentration, iron concentration, silica concentration and Cl-, and SO24 ions [1,31,32]. Cl-, and
SO24 are aggressive ions that interfere with the protective film and enhance localized corrosion.
Insoluble ferric ions (Fe3+) can form porous layers of rust resulting in fouling. Silicon dioxide
(SiO2), or silica, can form a tenacious silica scale that is hard to remove from the system. Organic
matter can foul the system, enhance the growth microbiology and cause under deposit corrosion.
Waters from which Ca2+ and Mg2+ have been removed result in different corrosion behavior than
waters that still contain hardness. Normally, water that has sufficient hardness and alkalinity will
result in CaCO3 precipitation at the cathodes. This layer of scale acts as a barrier against the
diffusion of oxygen and against electron transfer. On the other hand, it has been established that
bicarbonate in the absence of Ca2+ can act as an inhibitor to corrosion of steel, as contrasted to the
accelerative effect of Cl- and SO24 [32–34]. Indices have been developed that use the parameters
22
mentioned at the beginning of this section to provide information on the scaling and corrosion
tendency, which will be discussed in section 2.3.1.
Besides the water quality parameters, the effect of liquid velocity on the corrosion rate is also
important to take into account. According to several studies, the critical velocity above which the
corrosion rate decreases, because of the supply of oxygen for the formation of a passive film, is
about 0.7 m/s [35,36]. On the other hand, high water velocities can prevent the formation of
passive films through the action of shear stress.
2.2.4 Corrosion inhibitors

Treatments against electrochemical corrosion involves eliminating at least one of the four
requirements of the electrochemical cell, described in Section 2.2.1. Corrosion inhibitors are
chemical substances that are used to reduce the corrosion rate. Corrosion inhibitors can have three
common actions on an iron surface. First, the inhibitor can directly oxidize the ferrous ions to a
stable ferric oxide near the metal surface. Second, the inhibitor can catalyze the reaction of oxygen
with the ferrous ions. Third, the inhibitor can adsorb onto the surface to from a barrier or adsorb
onto the oxide layer to strengthen the protective film.
Another way of categorizing corrosion inhibitors is by subdividing the inhibitors in three

categories, namely anodic inhibitors, cathodic inhibitors, and film-forming inhibitors. Anodic
inhibitors interfere with the anodic reaction and are applied to prevent dissolution of the metal and
to prevent aggressive ions from reaching the metal surface. Although anodic inhibitors enhance
the formation of a protective oxide layer, a minor defect in the oxide film creates small anodic
sites that suffer from high corrosion rates, which results in pitting. Cathodic inhibitors interfere
with the cathodic reaction and are applied to prevent electron transfer at the cathode and the
reduction of oxygen. Mixed inhibitors, which are mostly organic inhibitors, provide a protection
mechanism that relies on the adsorption of the inhibitors, this can be either physical adsorption,
chemisorption or film formation. Chemisorption is more effective than physical adsorption
because it is non-reversible in many cases [37]. Adsorption relies on the structure of the inhibitor,
the surface charge of the metal, and on the type of electrolyte [38]. Just like in the case of anodic
inhibitors, minor defects can result in high corrosion rates.
23
Figure 7: Classification of inhibitors (based on diagram in Uhlig’s Corrosion Handbook [38])
Most corrosion inhibition programs rely on more than one type of inhibitor. A combination of
different inhibitors can reduce the total dose required. Effective inhibitors for SACW could
contain molybdate, zinc, amines or phosphonates. These inhibitors can be toxic at the required
concentrations, while phosphorus and nitrate containing inhibitors stimulate undesired
microbiological growth in the CWS and cause eutrophication of the environment. The debate on
silica as corrosion inhibitor is mature but still ongoing, since the silica chemistry is complex and
different for each type of water.
The selection of inhibitors for the corrosion experiments in this research was based on the criterion
of non-toxicity for the environment and on the availability for the experiments. The first selected
inhibitor was the organic film forming inhibitor created by Novochem Water Treatment called
NovoTraqua® (NT), and the second inhibitor was Na2SiO3 · 5H2O of Sigma Aldrich (≥95.0%).
Additionally, a Conventional All Organics (CAO) inhibitor was used, as this type of inhibitor is
common for alkaline environments.
Organic inhibitors often contain one or more of the heteroatoms, N, O, S, P, or Se, through which
the inhibitors anchor to the metal surface [38]. These functional groups act as an adsorption center
in the inhibitor molecule and adsorbs to the metal surface. NT is an organic inhibitor that relies on
both physical adsorption and chemisorption. It protects the metal by blocking and by hydrophobic
effects. The heteroatom of this inhibitor is O and the anchoring functional group is – COO-. The
polar heads of the surfactants orientate towards the metal surface, resulting in outward facing
hydrophobic groups that repel the cooling water An important benefit of NT is its biodegradability
and its non-toxicity. The standard NT product contains phosphonates at a concentration of 7.08%
total-PO34 .
All-organic treatment programs frequently incorporate phosphonates, which are very stable and
can operate at high alkalinities [39,40]. Phosphonates can act as both anodic inhibitor, by
catalyzing the oxidation of Fe(OH)2, and as cathodic inhibitor, if calcium or magnesium is present,
24
to form a scale barrier film. So the efficiency of corrosion inhibitors containing phosphonates in
SACW cannot rely on the cathodic inhibitor function.
Sodium silicate, or waterglass, is a corrosion inhibitor for soft waters that is appropriate in terms
of non-toxicity, non-phosphate, cheap and available [41]. Silica provides protection to iron by
forming an amorphous silica film which prevents the iron from dissolution. Lehrman and
Shuldener [42] state that solid corrosion products of the metal must be present before silica is
taken up from the solution. The negatively charged siliceous species react with the positively
charged iron ions. Despite its application, the inhibitive effects and mechanisms of sodium silicate
as corrosion inhibitor are not well understood and research is ongoing [40].
2.3 Corrosion measurements
2.3.1 Corrosion indices

No general corrosion index for predicting corrosion rates is available although several attempts
have been made. The developed indices often apply to specific cases only or are based on the
assumption that CaCO3 scale formation, as a protective layer on the metal surface, is possible. In
the case of soft water, where the Ca2+ and the Mg2+ ions are removed, the corrosion indices based
on scale formation are not suitable. This is also true for one of the most common saturation indices,
the Langelier index [8].
There are a number of water quality parameters that must be considered in the development of a
corrosion index or a series of indices. These include: hardness, alkalinity, carbonate, carbon
dioxide, pH, chloride, sulfate, ionic strength, conductivity, TDS, color, hydrogen sulfide, buffer
capacity, phosphate, silicate, DO, chlorine (free or combined), and temperature [7]. The corrosion
index must be designed so that the protective mechanism for a given water would predominate in
the calculation of the index.
Two corrosion indices that may be used as a rough estimate for the corrosion susceptibility are the
Larson-Skold index [9] and the Riddick index [10], which describe the corrosivity of water on
empirical bases. The Larson-Skold index (LSkI) is established on an empirical relationship
between polarization tests and mass loss determinations and is confirmed by Stern and Geary [43].
The study by Larson and Skold revealed a corrosivity relationship between the inhibiting property
of bicarbonate and the chloride and sulfate concentration, given in Equation (2.13) where all
concentrations are expressed as mg CaCO3/L.
Cl     SO42 
LSkI      (2.13)
 HCO3   CO3  
 2
   
25
The index can be interpreted by the guidelines given in Table 1.
Table 1: Interpretation guidelines for the LSkI

LSkI Indication
< 0.8 chlorides and sulfate probably will not interfere with natural
film formation
0.8 - 1.2 chlorides and sulfates may interfere with natural film
formation and higher than desired corrosion rates might be
anticipated
> 1.2 the tendency towards high corrosion rates of a local type
should be expected as the index increases
The Riddick Index is also an empirically developed index but takes more factors into account that
contribute to or that can mitigate the corrosion. The values obtained applied well to the soft waters
of the eastern seaboard of the United States but not to harder waters of the middle part of the
country [7]. The total hardness is defined by the calcium (Ca2+) and the magnesium (Mg2+)
concentration as equivalent parts per million CaCO3.
75   10   DO  2 
RI  CO2  12 ( Hardness  Alk )  Cl   2 N   (2.14)
Alk   SiO   Sat.DO 
 2 
Where
CO2 Free CO2 Concentration (mg/L)

Hardness Hardness (mg/L as CaCO3)
Alk Alkalinity (mg/L as CaCO3)
Cl- Concentration (mg/L)
SiO2 Concentration (mg/L)
N Concentration (mg/L)
DO Dissolved oxygen (mg/L)
Sat.DO Oxygen saturation at atmospheric pressure (mg/L)
Table 2: Interpretation guidelines for the RI
RI Indication
0-25 non-corrosive
26-50 moderately corrosive
51-75 corrosive
>75 very corrosive
Pisigan and Singley [32] proposed to combine the composite effect of pH and alkalinity into one
parameter, the buffer capacity  . The attempt was made to relate the buffer capacity to the
corrosion rate, but no definable relationship was found.
26
2.3.2 Mass loss measurements

The average corrosion rate over a specific period of time can be calculated by measuring the mass
loss of the specimen before and after immersing it in a corrosive environment and is calculated
with Equation (2.15). The corrosion products need to be removed by etching before weighing. The
duration of the tests depends on purpose of the test. Common testing periods for laboratory
immersion corrosion testing of metals is 1 to 10 days. [44]
KW
R (2.15)
ATd
Where:
K = a constant for calculating the corrosion rate in the right units;

= 3.45  106 for mpy
= 8.76  104 for mm/y
W = mass loss in g, to the nearest 1 mg (corrected for any loss during cleaning)
A = area in cm2 to the nearest 0.01 cm2
T = time of exposure in hours to the nearest 0.01 h;
d = density in g/cm3 which is 7.86 g/cm3 for carbon steel.
Corrosion inhibitors are generally applied to decrease the corrosion rate. The efficiency of the
corrosion inhibitors is calculated with Equation (2.16).
w0  w
IE  (2.16)
w0
Where w0 is the corrosion rate in the absence of inhibitor and w is the corrosion rate in the same
environment in the presence of the corrosion inhibitor. The corrosion rates can be expressed in
mils per year (mpy) or millimeter per year (mm/y) for example.
2.3.3 Voltammetry
Voltametry is a category of electroanalytical methods that can obtain information on the corrosion
behavior of a metal sample by controlling the potential and measuring the resulting current density
in a corrosion cell comprising of two or three electrodes. The potential of a system has to be
measured to a certain reference, which requires a reference electrode (RE). One type of
conventional reference electrode is the Saturated Calomel Electrode (SCE) that has a reference
potential of +0.241 relative to the Standard Hydrogen Electrode (SHE). The corrosion potential of
a system is determined by measuring the open circuit potential (OCP) that exists between the
metal sample, which is the working electrode (WE), and the reference electrode. In
potentiodynamic polarization tests, the potential is changed in a controlled manner using the WE
and RE, while the resulting current density is monitored using the WE and an auxiliary electrode
(AE). Common materials for auxiliary electrodes are graphite and platinum.
27
Current density versus potential plots are established by polarizing the WE by an external source,
the potentiostat. The obtained anodic and cathodic polarization curves establish the Tafel-plot of
which an example is given in Figure 8. The corrosion potential and current can be determined at
the intersection of the linear derivative of the anodic and cathodic branch in the Tafel-plot because
the corrosion rate depends on the equilibrium kinetics of the oxidation and reduction reactions.
Faraday’s law describes the linear relationship between the corrosion rate R (or metal dissolution
rate) and the corrosion current icorr as is given in (2.17).
M
R icorr (2.17)
nF 
Where M is the atomic weight of the metal, n is the charge number, F is the Faraday constant, and
ρ is the metal density.
The resulting polarization curves show the corrosion behavior under cathodic or anodic
polarization, and indicate the regions of immune, active, passive, transpassive behavior, as is
shown in Figure 9.
Cyclic voltammetry is a cyclic potentiodynamic polarization (CPP) test in which the potential is
reversed after reaching a potential or current density setpoint. CPP tests are performed to evaluate
the susceptibility to pitting initiation and propagation. A large increase in the current density marks
the initiation and propagation of localized corrosion, which is called the breakdown point. The
potential is reversed after the local breakdown of the passivity. If the specimen is able to
repassivate after the breakdown, the current density will also decrease again. If the specimen is
not able to repassivate, the current density will remain high with decreasing potential. Increasing
the potential without breakdown of the passivity, oxygen evolution as in Equation (2.8) can be
observed. The possible CPP curves are described in Figure 10, where Eocp is the OCP, Eb is the
breakdown potential and Ep is the repassivation potential.
28
Figure 8: Hypothetical cathodic and anodic polarization diagram (ASTM G3 [45]).
Figure 9: Hypothetical cathodic and anodic polarization plots for a passive anode (ASTM G3 [45]).
29
Figure 10: Hypothetical CPP diagrams where a) the sample is repassivated after local passivity breakdown, b) the
sample is not repassivated after local passivity breakdown, and c) the breakdown point is not reached and oxygen
evolution occurs.
30
2.3.4 Surface analysis

Different surface analyses techniques are useful for a better understanding of corrosion behavior
and for verification of results obtained by mass loss tests and electroanalytical methods. Scanning
electron microscopy (SEM) in combination with energy dispersive spectroscopy (EDS) is a
common method in corrosion research used to study the surface morphology and to determine the
composition of the top layer of the specimens on a microscale. SEM/EDS collects information of
the sample its surface by scanning the surface with a focused beam of electrons in a vacuum
chamber. The beam of electrons excites the atoms of the specimens surface resulting in a
characteristic backscatter of electrons released by the specimen, which gets translated to an image
or elemental analysis based on the energy that is transmitted by the electrons of the exited atoms.
SEM could create images from nanosized areas because this surface analysis technique does not
rely on the magnifying power of an objective lens. While SEM is used to create images, EDS is
used for the elemental analysis of the surface of a sample. (E. Reinton, personal communication,
March 1, 2018)
Another advanced surface analysis method is X-ray diffraction (XRD). This method can determine
the crystalline structure of the top layer of a specimen. A beam of send X-rays gets diffracted by
the top layer of the specimen and the angles and intensities of the reflected beams are used to
identify the crystalline structure. Amorphous layers will not diffract the X-rays in an identifiable
pattern. (R. Hendrix, personal communication, September 27, 2017)
31
3 Synthetic Cooling Water
The lab experiments that tested the corrosion behavior of mild steel in CWS were executed in
water that resembled the main characteristics cooling water. Three different types of cooling water
were considered for the experiments, actual cooling water from on-site, lab-concentrated surface
water, and synthetic cooling water. The first option, actual cooling water, could have been
obtained from a site in the Botlek operated by Evides Industrial Water. The second option,
concentrated surface water, could have been obtained by applying small scale treatment steps to
surface water, followed by a concentration step by means of a reverse osmosis (RO) membrane.
The surface water was available from a pipeline that transported water to the laboratory from a
nearby major water canal. The surface water would have been concentrated by recirculating the
concentrate stream from the RO membrane to reach the desired concentration ratio. The third
option, synthetic cooling water (SCW), comprised ultrapure water as a base matrix and a well-
defined dosage of different salts.
The two key criteria for selecting the type of cooling water that would be used in lab experiments
were repeatability and representability. An additional criterion was self-sufficiency. Out of the
three options, SCW was selected based on these criteria. This chapter describes how the SCW was
produced, what the characteristics were and how the calculations compared with the results
calculated by the software PHREEQC with the proposed synthetic water composition as input file.
3.1 Materials and Methods
3.1.1 Synthetic cooling water production

The composition of the SCW was derived from the composition of Dutch surface water. Table 3
gives an overview of the main water constituents of three surface waters in the area that are
indirectly used for cooling water. Based on the values of these three water bodies, values for
synthetic surface water were selected. The COC determined with what ratio the ion concentrations
of the SSW must increase to obtain SCW. Depending on what COC was selected the desired ion
concentrations and therefore the required salt dosages could be calculated. The procedure for
making SCW was based on the Dutch protocol for preparing synthetic laboratory samples [46], a
guidance on metals in aqueous media [47], and a standard for the preparation of substitute ocean
water [48].
33
Synthetic Cooling Water
Table 3: Main water composition of the river Meuse at Keizersveer (RIWA), the surface water reservoir Peterusplaat
(Evides), the surface water reservoir Braakman (Evides) and the selected guideline for the SSW composition.
Keizersveer Petrusplaat Braakman SSW
pH (-) 7.94 8.53 8.02 -
Temp (°C) 13.2 12.4 13.2 -
EC (µS/cm at 20°C) 472 419 572 564
TOC (mg/L) 4.8 4.1 8 5.1
HCO3- (mg/L) 169 126 194 145
Na+ (mg/L) 31.5 32 38 32
K+ (mg/L) 6.6 6.4 12 8
Ca2+ (mg/L) 58.1 49 74 60
Mg2+ (mg/L) - 7.3 10 8
NH4+ (mg/L) 0.145 0.06 0.03 0.1
Cl- (mg/L) 44.8 45.5 66.4 46
SO42- (mg/L) 50.3 50 56 52
NO3- (mg/L) 14.5 10.3 13.7 13
P-tot (mg/L) 0.39 0.3 0.013 0.35
Fe-tot (mg/L) 0.61 0.013 0.04 0.1
Mn (mg/L) 0.674 0.01 0.01 0.01
F (mg/L) 0.229 0.21 0.2 0.21
Br (mg/L) 0.0935 0.08 - 0.09
SiO2 (mg/L) 2.56 2.1 1.7 2.2
The total organic carbon (TOC) in the SCW was simulated by dosing humic substances.
Commercial humic substances, like sodium humic acids of the brand Sigma Aldrich, are
considered inappropriate for the simulation of surface water [49]. As an alternative, the product
HumVi was used. This is a concentrated and desalinized product from the regenerate stream of
the ion exchange process in the drinking water production line of Spannenburg, Vitens N.V. and
is high in fulvic acids. Fulvic acids are the main form of humic substances in surface waters [50].
Additional information on HumVi is provided in Appendix A. Although the remaining salt
concentration of HumVi is low, it is accounted for in the calculation of the required salt dosages
of the SCW.
The different ions in the SCW can precipitate under certain conditions. Iron, manganese, calcium
and magnesium precipitate at a high pH, whereas the solubility of silica in water increases above
pH 8 [51]. In case SCW with a COC of 10 is prepared containing all components listed in Table
3, multiple stock solutions have to be prepared that favor the solubility of the different
components. These stock solutions are at least a tenfold more concentrated than the SCW that is
used in the experiments. This means that when a COC of 10 is used in the experiments, the stock
solutions must be at least a COC of 100. These solutions are prepared with high purity chemicals
and ultrapure water, as is summarized in Table 4.
34
Table 4: Chemicals used to create the stock solutions to produce SCW.

Brand Purity Molar mass
(-) (%) (g/mol)
Stock sol. A
Na2SO4 Sigma Aldrich ≥ 99.0 142.04
NaCl VWR ≥ 99 58.44
NaNO3 Sigma Aldrich ≥ 99.0 85.00
KCl Sigma Aldrich 99.0-100.5 74.55
Na3PO4 Sigma Aldrich 96 163.94
NaF Sigma Aldrich ≥ 99 41.99
NH4HCO3 Sigma Aldrich ≥ 99.5 79.06
NaBr Sigma Aldrich ≥ 99.5 102.89
HumVi Vitens See App. A -
Stock sol. B
Na2SiO3 · 5H2O Sigma Aldrich ≥ 95.0 212.14
NaOH Sigma Aldrich ≥ 97.0 40.00
Stock sol. C
FeCl3 · 6H2O Sigma Aldrich ≥ 98 270.30
MnSO4 · H2O Sigma Aldrich 99.0-101.0 169.02
CuSO4 Sigma Aldrich 98 159.61
H2SO4 Sigma Aldrich 95.0-98.0 98.09
Three stock solutions were proposed:
Stock A: Bulk solution with soluble salts;

Stock B: Silicate solution with raised pH;
Stock C: Transition metals solution with lowered pH.
The preparation of the stock solutions was done in volumetric flasks. The volumetric flasks were
acid-cleaned and rinsed with demineralized water. The mass of the chemicals was determined on
a mass balance with a minimum display of 0.01 mg. The chemicals were dissolved in the
volumetric flasks that was filled for 90% with ultrapure water. The pH of the water for stock
solution B and C was adjusted, with NaOH and H2So4 respectively, before the other chemicals
were dosed. After the last dosage, the flask was filled up with ultrapure water to reach the 100%
of the required volume. The solution was mechanically mixed by a magnetic stirring rod during
the dosages of the chemicals and another 24 hours afterwards. The stock solution was stored in a
glass bottle that was closed to the surrounding air, in the dark, at approximately 4°C until use, and
during a maximum of two months.
The three stock solutions were diluted and mixed upon usage to the desired COC. The mixture
was left to age for 24 hours under continuous mixing and was used for a maximum of 1.5 weeks
in the experiments. The SCW mixture required aeration with air from the atmosphere to ensure
that the concentration of dissolved oxygen reached 100% saturation and that a CO2 equilibrium of
an open CO2-system was reached.
35
The stock solution and the mixture were analyzed for pH, electrical conductivity (EC) and
concentration of the components. Ion chromatography was used to measure the concentration of
Ca 2+ , Mg2+ , Na + , K+ , NH4 , F , Cl , Br , NO2 , NO3 , SO42 and PO34 by means of the 883
Basic IC Plus in combination with the 881 Compact IC pro of the manufacturer Metrohm. Total
carbon and inorganic carbon measurements were used to determine the TOC by means of the
TOC-V CPH in combination with the ASI-V of the manufacturer Shimadzu. The iron and copper
concentrations were measured using Merck Millipore test kits and the silica concentration was
measured using Hach Lange test kit.
The calculations of the EC were based the Nernst-Einstein equation that relates the molar limiting
conductivity and the diffusion coefficient for a given ion. The method of Appelo [52], who adapted
this equation, was used to calculated the EC. The EC-values of the water at different temperatures
were corrected using linear Taylor expression obtained from the non-linear compensation model
and the parameters of Atkins [53], resulting in the following more simple equation:
ECT
EC25  (3.1)
(1  0.020(T  25))
The aqueous solution must be electrically neutral. Because the SCW was based on surface water
composition reports of which the reported values may have an electrical imbalance, the resulting
SCW may also have an electrical imbalance. The charge-balance error (CBE) is calculated
according to Equation (3.2) and a CBE of 5% or less is acceptable [54].
CBE 
 cations   anions 100 (3.2)
 cations   anions
3.1.2 PHREEQC
The water in a CWS is in contact with the atmosphere and changes temperature when passing the
heat exchanger or during the evaporation process. The water quality is therefore rather dynamic.
The SCW used in the lab experiments to test corrosion was under constant conditions. The
outcomes of the PHREEQC model helped to verify whether a sound composition was selected for
the concentrated SCW, whether the calculations were correct and what kind of precipitation
products were to be expected. The precipitation products are based on the equilibrium phases
specified in the PHREEQC database.
The pH is initially provided in the input file but is corrected by the partial pressure of CO 2 for an
open system. The CO2 concentration of the water exiting the piping system and entering the free
atmosphere of the cooling tower is 800 ppm [55]. The partial pressure used in the input file is
therefore 10-3.1 atm. The evaporation or the increase in COC is simulated by removing stepwise
moles of water from the solution. The PHREEQC input file can be found in Appendix B.
36
3.2 Results and Discussion
3.2.1 SCW composition

It was decided that the SCW would resemble cooling water with a COC of 10. Advanced makeup
water treatment would only be economically viable if high COCs are possible. The high ion
concentrations and the resulting high pH level of cooling water with a COC of 10 would make it
interesting to research the corrosion behavior and seek for alternative treatment opportunities.
Some simplifications were made in the final composition of the SCW. In practice, strong cation
exchange (CIEX) could be applied to remove the hardness. For the composition of the SCW, it
was assumed that Ca2+ and Mg2+ were completely exchanged by Na+ although CIEX does not have
a 100% exchange efficiency. The concentrations of Cu2+ and Mn2+ are very low and would
normally decrease even more after CIEX. Therefore, these ions were removed from the water
composition.
Another simplification was the removal of HumVi from the initial SCW composition. TOC can
increase the solubility of compounds that would have precipitated without the presence of TOC.
The corrosion behavior of mild steel is also influenced by humic acids without knowing exactly
by what mechanism and under which conditions. Therefore it was decided that HumVi is not
included in the production of SCW.
The last component that was removed from the original composition was Fe3+. Fe3+ ions precipitate
at the natural pH of the SCW. Electrochemical measurements would have been undesirably
affected by the precipitation products.
These simplifications resulted in a far easier procedure to produce SCW. First of all, a bucket was
filled with 20 L ultrapure water. Subsequently the different chemicals and their dosages, as
summed up in Table 6, were put in a small container. At last the dry mix of chemicals was slowly
added to the water while the water was continuously stirred.
The final composition of the SCW is presented in Table 5 and the dosages to make this
composition is given in Table 6. The carbonate equilibrium was selected at pH 9.5 and at 20°C,
which resulted in 0.1% H 2 CO3 , 88.3% HCO3 and 11.6% CO32 based on the equilibrium equations
of Harned and coauthors [56,57]. The CBE for this composition is -0.36% which is a good
equilibrium of the charges.
37
Table 5: SCW composition with a COC of 10. Table 6: Dosages for SCW with a COC of 10.
pH (-) 9.1  9.4 Dosage
Temp (°C) 20 (mg/L)
EC (mS/cm) 4.4  4.5 Na2SO4 769
TOC (mg/L) 0 NaCl 639
HCO3- (mg/L) 1251 NaNO3 178
Na+ (mg/L) 1166 KCl 153
K+ (mg/L) 80 Na2SiO3 · 5H2O 73
Ca2+ (mg/L) 0 Na3PO4 6.0
Mg2+ (mg/L) 0 FeCl3 · 6H2O 0
NH4+ (mg/L) 1.0 NaF 4.4
Cl- (mg/L) 460 NH4HCO3 4.4
SO42- (mg/L) 520 NaBr 1.2
NO3- (mg/L) 130 NaHCO3 1722 1601
P-tot (mg/L) 3.5 Na2CO3 292  366
Fe-tot (mg/L) 0
Mn2+ (mg/L) 0
F- (mg/L) 2.0
Br- (mg/L) 0.9
SiO2 (mg/L) 20
3.2.2 PHREEQC
Actual mixtures, according to the dosages of Table 6, resulted in a pH of 9.1 and an EC of 4.4
mS/cm. These values were in line with the results from PHREEQC. The input values for the
solution and the eventual solution according to this model are given in Figure 11. Since the desired
pH for the corrosion experiments was 9.4, the NaHCO3 and the Na2CO3 dosages were adjusted to
1601 mg/L and 366 mg/L respectively.
Simulating the effect of higher COCs on the saturation indexes and the pH in PHREEQC resulted
in the graphs of Figure 12. The model gives an initial idea of possible scale formation that were
to be expected in experiments with the SCW. Ca2+ and Mg2+ obviously cause large scaling
problems. Softening the water would reduce the risk of scaling. However, silica can cause a
persistent scale in both hard and soft water and it can form complexes with iron. Quartz and other
crystalline silica polymorphs exhibit slow precipitation kinetics, while amorphous silica or
silicates tend to precipitate rapidly as scale deposits via condensation polymerization [58]. An
elevated pH can increase the solubility and diminish the risk of scaling. Silica and silicates can
also be considered as corrosion inhibitors. This is a short and incomplete summation of the effects
of silica and silicates. Further research on the risks and benefits of silica in SACW is required,
especially on the precipitation of metal-rich amorphous silica and at elevated temperatures.
38
Figure 11: PHREEQC input solution (left) and output characteristics (right).
Figure 12: Impact of an increasing COC of SCW on the Saturation Index (SI) and the pH
39
3.2.3 Corrosion indices

The LSkI and the RI were calculated for the surface waters mentioned in Table 3, for the SSW
and for the SCW. All waters were classified as moderately corrosive according to the LSkI and
the COC of the water did not influence the value as was expected from evaluating the equation.
An increase in the COC must however lead to an increase in the corrosion susceptibility as is
experienced in practice.
The values according to the RI showed great variation that could be linked to the hardness
concentration and alkalinity. The weight of ‘hardness – alkalinity’ in equation is too large for the
75 1
surface waters containing some hardness, even with the applied factor of  . Singley [7]
Alk 2
already stated that the index is not suitable for hard waters. In addition to this, it can be concluded
that the RI is neither suitable for corrosion estimations of mild steel in Dutch surface waters. The
RI for the SSW was classified as moderately corrosive while the RI for SCW was classified as
non-corrosive. In this case mainly due to the high alkalinity of SCW.
DO  2
The element in the index is not suitable for CWS since the water is oxygen saturated in
Sat.DO
these systems and oversaturated when heated at the heat exchangers. The absence of oxygen is
beneficial for corrosion reduction, yet it should not be considered in this manner. The aim of
SACW is natural passivation which requires oxygen. Therefore oxygen should have a beneficial
effect on the corrosion reduction, and only if the water is oxygen saturated. Another element that
is missing in the RI is the effect of SO42 .
Although the limited application of corrosion indices was already stressed, it can be concluded
that the corrosion indices that appeared to be most suitable for SACW of all corrosion indices, are
unsuitable for corrosion estimations in this type of water.
40
4 Mass loss tests
Laboratory immersion corrosion tests were carried out to gain fundamental knowledge on the
corrosion of mild steel in the SCW and the effect of different inhibitors on this corrosion process.
Mass loss tests are the most common of all corrosion-test measurements for determining corrosion
rates [59] and are widely applied in the field. Combining polarization resistance tests with mass
loss tests can give useful insights. Mass loss measurements provide information on the average
corrosion rate during a certain period, while polarization resistance tests provide information on
the change in the corrosion rate during this period. This chapter will focus on mass loss tests and
the subsequent surface analyses.
Two types of mass loss tests were defined in which the mass loss of metal coupons were
determined; 48 hours immersion and 4 weeks immersion. The 48 hours tests were to provide an
initial understanding on the corrosion behavior and are an uncomplicated method to generate an
overview on the inhibition efficiency of different inhibitors, concentrations and conditions. Mass
loss test should however run over longer periods of time to understand the development of the
corrosion rate. Therefore a 4 weeks immersion test was designed that would provide information
on the average corrosion rate per week during a period of four weeks.
This chapter describes first the coupon composition and coupon preparation, followed by a
description of the setup, the test solutions, the surface analysis method and at last an overview of
the procedures. The referenced standards that have been used as guidelines for creating an
laboratory immersion test setup were:
G1 Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens
[60];
G3 Conventions Applicable to Electrochemical Measurements in Corrosion Testing
[45];
G5 Standard Reference Test Method for Making Potentiodynamic Anodic Polarization
Measurements [61];
TM0169/G31 Standard guide for laboratory immersion corrosion testing of metals [44].
4.1.1 Coupon preparation and composition

The 48 hours immersion tests involved coupons that were made from cold rolled steel with the
following composition (wt.%): 0.087 C, 0.375 Mn, 0.013 P, 0.013 S, 0.018 Si, and remainder Fe.
The specimens had the dimensions of 80  10  1.5 mm. The steel surface was prepared by
mechanical wet grinding and polishing with silicon carbide paper of 320 to 1200-grit.
Subsequently, the coupons were ultrasonically degreased with acetone for 15 minutes, rinsed with
ethanol, rinsed with distilled water and dried in air. The mass of all the coupons was determined
41
Mass loss tests
on an analytical balance to the nearest 0.1 mg before the coupons were partially isolated from the
environment with 3M™ polyester tape 8402 with silicone adhesive, leaving three circular areas
of in total an area of 2.36 cm2 to react with the environment. The adhesive has a high chemical
resistance, also in anodizing environments [62], and is therefore suitable for the immersion tests.
The specimens were finally stored in a desiccator and used for an experiment within one hour after
storage.
The 4 weeks immersion tests involved standardized flat coupons with stamped identification
numbers from Metal Samples Co with the chemical composition (wt.%): 0.08 C, 0.330 Mn, 0.011
P, 0.009 S, 0.045 Al, 0.019 Si, and remainder Fe. The dimensions of the coupons were 72  94
 1.3 mm. The coupons were cleaned according to the same procedure as described for the 48
hours tests. No polishing was required (Novochem and Nalco, personal communication, 2017).
For the 4 weeks immersion tests, the top 26 mm of the coupons, including the stamp, were isolated
with the polyester tape, leaving an area of 46  9.4  1.3 mm to react with the environment. The
dimensions of the exposed areas were determined with an accuracy of at least ±1% by using a
digital Vernier caliper. Underneath the polyester tape, copper tape was applied to ensure a
conducting connection between the sample and the potentiostat for polarization tests.
After the required immersion time and corrosion product analyses, the coupons were etched in an
inhibited 3% HCl solution. The etching removed the corrosion product to unveil the metal surface
and to determine the actual mass loss, again on an analytical balance to the nearest 0.1 mg.
4.1.2 Setup
The 48 hours tests implied a simple setup of acrylic containers with 210 mL of solution that was
continuously stirred during 48 hours with a magnetic stirring bar. The metal surface of the coupon
coupon was immersed in the solution in such a way that the rim of the polyester tape hung 1 cm
under the water surface.
The setup for the 4 weeks immersion test consisted of six 2L polypropylene measuring jug in
which four coupons could be immersed. A schematization and a photo of the setup are given in
Figure 13 and Figure 14, respectively. The jugs were filled with a 1600 mL solution that was
continuously refreshed with SCW that was pumped from a 10L bucket into the jug by a 205S
Watson Marlow peristaltic pump. The pump had a capacity of twelve channels, of which six were
assigned for the influent flow and six were assigned for the effluent flow. The effluent was pumped
from the jugs into separate 2L polypropylene bottles. The hydraulic retention time was four days.
The influent was air saturated at 1 atm by aerating the influent SCW with an aeration stone and
stirring the water with a magnetic stir bar. All jugs were stirred with by the JLT6 jar test of Velp
Scientifica by means of a mechanical mixer that created an average velocity of 1 m/s at the outer
edges of the jugs. The minimum ratio of test solution volume to test specimen surface area of 0.20
mL/mm2 was recommended by ASTM G31 [44]. With a total surface area of almost 1000 mm2
for the four coupons and a solution volume of 1600 mL, this recommendation was met.
42
Mass loss tests
The measuring jugs were closed off with custom made lids, which included coupon holders,
openings to mount sensors, and an opening for the mechanical stirring rod. The spout was closed
off with plastic paraffin film. All material, other than the coupons, was made of plastic if possible.
This was to avoid glass contamination. The lids were designed in such a way that they would also
fit on the original glass beakers belonging to the JLT6 jar test. A photo of the jug and lid are given
in Figure 15 and Figure 16. The coupons were mounted to the coupon holders with a plastic screw
and nut, 3 cm above the water surface to avoid crevice corrosion. The length of the coupon holders
ensured that the coupons were covered with polyester tape at the height of the water-atmosphere
interface, to avoid water line corrosion. The coupon holders were located on the outer edges of the
lid while the sensors were located in the center of the lid.
The sensors measured the pH, EC and the temperature every hour. The water temperature was
manually regulated with a six position heating plate. Initial tests indicated that the temperatures
were not sufficiently constant since they could vary ± 4°C among eachother. Therefore, the final
tests were run at a room temperature of 21.0 °C ± 1.5°C and the heating plates were no longer
used.
Figure 13: Schematic representation of the 4 weeks immersion test setup of a single jug.
(See Appendix C for the multiple jug setup schematization.)
43
Mass loss tests
Figure 14: Photo of the 4 weeks immersion test setup.
Figure 15: 2L Measuring jug with custom made lid. Figure 16: Custom made lid with fixed positions for
coupon holders, sensors, stirring rod and influent and
effluent syringes.
4.1.3 Test conditions

The experiments of the 48 hours immersion were started with four different solutions. Depending
on the results, more solutions would be tested and all tests would be executed in triplicates. The
four solutions that were tested in duplicates at first were:
Blanc: only SCW;

CAO: SCW with 200 ppm Conventional All Organics inhibitor;
NT100: SCW with 200 ppm NT containing 100% of the original phosphorus concentration;
Si200: SCW with 200 ppm SiO2 dosed as Na2SiO3 · 5H2O.
44
Mass loss tests
These tests would have been followed up by immersion test with the following composition:
NT12: SCW with 200 ppm NT containing 12.5% of the original phosphorus concentration;
Si300: SCW with 300 ppm SiO2 dosed as Na2SiO3 · 5H2O;
The solutions of the 48 hours experiments where not refreshed during those 48 hours. This was
different for the solutions in the jugs of the 4 weeks immersion tests. Those solutions were
continuously refreshed based on a hydraulic retention time of four days. The setup provided the
possibility of running experiments in six jars simultaneously. Therefore, four different solutions
were tested in triplicates during two months, namely Blanc, CAO, NT100 and Si200. The dosages
of the inhibitors were calculated based on the HTI. The HTI of the setup, with a volume of 1600
mL and a hydraulic retention time of four days, was 2.8 days. Figure 17 shows the development
of the inhibitor concentration over time if the inhibitor concentration decreases only due to SCW
inflow and inhibitor outflow with the effluent. The inhibitors were dosed directly in the jugs with
a Thermo Scientific Finnpipette. The inhibitors were dosed with a 24 hour interval, therefor the
dosages were based on the desired concentration after 24 hours. This means that the required
dosage for a system with a volume of 1600 mL and a desired concentration of 200 ppm, was 528
μL . A double inhibitor dosage was applied upon startup, as is common in practice to ensure full
protection by the inhibitor.
Inhibitor concentration development based on HTI

1200
Inhibitor concentration (ppm)
Startup dosage (ul)

Daily dosage (ul)
900
600
300
0
-24 -12 0 12 24 36 48
Time (hours)
Figure 17: Inhibitor concentration development due to inflow of SCW and

outflow of inhibitor with the effluent.
45
Mass loss tests
4.1.4 Surface analyses

The surfaces of the coupons were photographed and evaluated before immersion, after immersion
when the corrosion products were still visible, and removal of the corrosion products by etching,
which revealed the surface attack. The surface attack was categorized in uniform corrosion,
crevice corrosion and pitting corrosion based on visual inspection with 15x magnification when
necessary. The pitted surfaces were rated in terms of pit density based on the ratings in Table 7.
Differentiation was made between the outward facing surface and the inward facing surface.
Vertical cross-sectioning was required to determine the pit type as presented by Figure 6 [30],
however this was not part of the scope.
Table 7: Rating table for pit density.

Pit density
(number of pits/m2)
P-1 <4  103
P-2 <2  104
P-3 <8  104
P-4 <2  105
P-5 >2  105
4.1.5 Procedure
The previous sections provide information on the coupon preparation, the setup, the test conditions
and the surface analyses. This section provides a stepwise overview of the actions for the mass
loss tests. The 48 hours immersion test was a simple test and its procedure is summed up below.
1. Coupon preparation as in Section 4.1.1;

2. Solution preparation as in Section 4.1.3;
3. The pH, EC and temperature of the solution were measured;
4. The coupon was immersed in the solution;
5. The coupon was removed from the solution after 48 hours;
6. The surface was photographed;
7. The tape was removed;
8. The surface was etched;
9. The mass of the coupon was determined;
10. The surface was photographed again;
11. Surface analyses was performed;
12. The coupons were numbered stored in small sealed bags.
The goal of the 4 weeks immersion test was to determine the average corrosion rate per week
during a period of four weeks based on mass loss, and to determine the corrosion rate after two
and four weeks. The results of the mass loss test and the polarization resistance tests were to be
compared. The procedure of the 4 weeks immersion test is summarized on the next page.
46
Mass loss tests
At the startup of the 4 weeks immersion test:

1. Setup preparation as in Section 4.1.2;
2. Coupon preparation as in Section 4.1.1;
3. Solution preparation as in Section 4.1.3;
4. The pH, EC and temperature of the solution were measured;
5. SCW was pumped through the entire setup;
6. 4 coupons per jug were mounted to the lid and immersed in the solution;
During the 4 weeks:

7. The pH, EC and temperature of the solution were measured ervery hour automatically and
manually every day;
8. One inhibitor dosage was applied every day around 3 PM;
9. One coupon per jug was replaced after 1 week and after 2 weeks of immersion;
10. The procedures 6 untill 12 of the 48 hour immersion test were followed for the removed
coupons;
11. All coupons were removed after running the setup for 4 weeks;
a) The coupons that were immersed for 2 and 4 weeks were eached moved to a
separate container with effluent from the correstponding jug and connected to the
potentiostat to run a polarization resistance test;
b) The coupons that were immersed for 3 and 4 weeks followed the procedures 6
untill 12 of the 48 hour immersion test;
At the end of the 4 weeks immersion test:

12. The setup is cleaned with acid and rinsed with distilled water.
All mass loss measurements were done in triplicates. Blanc, CAO and NT100 were tested during
the first run of four weeks with two jugs per solution. Si200, Blanc, CAO and NT100 were tested
during the second run. For this run, three jugs were assigned to the Si200 solution and the
remaining three jugs were assigned to the Blanc, CAO and NT100 solutions.
4.2.1 Mass loss measurements

The mass loss of mild steel coupons was studied in different solutions and in different set ups. No
visual corrosion of the coupons was observed in the 48 hours immersion tests. This can be related
to the protection of the mild steel by passivation, as will be further discussed in Chapter 5. The 4
weeks immersion tests did result in corrosion rates and inhibitor efficiencies, and will have the
main focus in this section.
The mass loss of mild steel coupons was determined every week during a period of four weeks.
Based on the mass loss measurements, the corrosion rates were calculated with Equation (2.15).
47
Mass loss tests
The resulting corrosion rates of the coupons in the different solutions are summarized in Figure
18. The error bars indicate a two-sigma confidence interval. A maximum corrosion rate of 2 mils
per year (mpy) is an acceptable corrosion rate of mild steel in full scale CWS. However, this is
not a justifiable maximum corrosion rate for laboratory immersion tests. The corrosivity of the
solutions to mild steel should be determined in relative, rather than absolute, terms [63].
Figure 18: Corrosion rates for Blanc, CAO, NT100 and Si200 at a weekly interval.
The solution without inhibitors, i.e. Blanc, had the highest corrosion rates, followed by CAO,
NT100 and Si200. This was conform the expectations because the addition of an inhibitor is
supposed to lower the corrosion rates. Additionally, CAO is not designed for soft water and
therefore provided only little protection against corrosion. The performance of NT100 was
unknown so far for soft water at a pH above 9.3. The result was that it cut the corrosion rate of the
Blanc solution by half. The inhibition efficiency of NT100 might have been higher if the
concentration of inhibitor was increased. With this knowledge, the voltamettric tests described in
the next chapter were carried out with higher concentrations. Si200 performed best of all inhibitors
and no silica scale was observed in the system. Si200 is therefore considered as an inhibitor that
is worthwhile to be further researched in soft alkaline cooling water conditions. The uncertainty
was too large to state that the corrosion rates of one of the solutions decreased or increased during
the four weeks period.
The inhibition efficiencies of CAO, NT100 and Si200 with respect to Blanc could be calculated
based on the corrosion rates with Equation (2.16). Again, the effect of the uncertainty of the
corrosion rates on the inhibition efficiency is too large to draw conclusions in terms of inhibition
efficiency.
48
Mass loss tests

Visual surface analyses were carried out to study the characteristics of the corrosion products
before etching and of the surface of the coupon after etching. The analysis focused on uniform,
crevice and pitting corrosion. It is possible that all three types of corrosion are found on a coupon.
A distinction was made between the outward facing side of the coupon (F) and the inward facing
side (B). Figure 19.a provides an overview of the number of coupons that exhibited uniform
corrosion and Figure 19.b of the number of coupons that exhibited crevice corrosion, respectively.
Figure 19.c provides an overview of the rating of the pit density based on the cut off values of
Table 7. Where P-5 is an indication for a high pit density and P-0 for a low pit density, respectively.
For each of the four solutions, twelve coupons were studied.
Although the uncertainty of the calculated corrosion rates was too large to state that the corrosion
rate of the coupons in the Blanc solution decreased, the tendency may be true and be explained by
the formation of a uniform protective layer of oxidation products. Mainly uniform corrosion and
little pitting corrosion was identified on the coupons of the Blanc solution. With the formation of
a uniform corrosion product on the steel surface, it is expected that the corrosion rate stabilizes
over time.
It is important to note that most mass loss results were biased by crevice corrosion. The coupons
in the Si200 solution suffered least from crevice corrosion, and also least from pitting corrosion.
Particularly the coupons in the CAO and NT100 solutions suffered from pitting corrosion and
partially or no uniform corrosion. Additionally, more uniform corrosion was observed on the
inward facing side of the coupons which is related to a difference in the flow along the metal
surface. Inserting coupons in a recirculating plug flow, like a corrosion rack, would diminish the
effect of different flows along the coupon surface. So the bias of crevice corrosion, the non-
corresponding types of corrosion among the coupons of the different solutions, and the difference
between the inward and outward facing sides, made it faulty to draw conclusions in terms of best
protection. It was right to conclude that calculating the inhibition efficiencies would be misleading.
Still, it can be concluded that the Si200 solution has the best inhibiting effect on mild steel
corrosion and further research on Si200 for this type of cooling water is recommended.
Electroanalytical methods in combination with surface examination were likely to provide more
insights on the performance of the corrosion inhibitors and enable comparison between the
different inhibitors. Since pitting was the main corrosion process in the solutions with an inhibitor,
CPP tests were selected to evaluate the performance of the inhibitors. CPP tests are discussed in
the next chapter.
49
Mass loss tests
Uniform Corrosion
14
12
10
0
Mainly uniform Partially uniform No uniform
Crevice Corrosion
14
12
Number of coupons
10
0
Substantial crevice Some crevice No crevice
Pitting Corrosion
8
7
6
5
4
3
2
1
0
P-5 P-4 P-3 P-2 P-1 P-0
Blanc.F Blanc.B CAO.F CAO.B NT100.F NT100.B Si200.F Si200.B
Figure 19: Results of surface analysis after immersion in terms of a) Uniform corrosion, b) Crevice corrosion, and
c) Pitting corrosion. Distinction was made between the front/outward facing side of the coupon (F) and the
back/inward facing side of the coupon (B).
50
5 Voltamettric tests
The behavior of mild steel in SCW in combination with different inhibitors and concentrations
has been studied by means of cyclic potentiodynamic anodic polarization measurements (CPP)
based on ASTM G61 [64]. This method provided an indication of the susceptibility to initiation
of localized corrosion. The more noble the potential at which the anodic current increases rapidly,
the less susceptible the alloy is to initiation of localized corrosion. The potential at which the
hysteresis loop is completed is another indication of the likeliness of localized corrosion to occur.
This chapter describes the experimental conditions and the results of the voltamettric tests in SCW.
The standards that have been used for the voltamettric test as reference material are:
F746 Standard Test Method for Pitting or Crevice Corrosion of Metallic Surgical
Implant Materials [65];
F2129 Standard Test Method for Conducting Cyclic Potentiodynamic Polarization
Measurements to Determine the Corrosion Susceptibility of Small Implant Devices [66];
G1 Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens
[60];
G3 Conventions Applicable to Electrochemical Measurements in Corrosion Testing
[45];
G5 Standard Reference Test Method for Making Potentiodynamic Anodic Polarization
Measurements [61];
G46 Standard Guide for Examination and Evaluation of Pitting Corrosion [30];
G61 Standard Test Method for Conducting Cyclic Potentiodynamic Polarization
Measurements for Localized Corrosion Susceptibility of Iron-, Nickel-, or Cobalt-Based
Alloys [64].
5.1.1 Coupon preparation and composition

The experiments were carried out with specimens made from cold rolled steel with the following
composition (wt.%): 0.087 C, 0.375 Mn, 0.013 P, 0.013 S, 0.018 Si, and remainder Fe. The
specimens had the dimensions of 22  22  1.5 mm. The steel surface was prepared by
mechanical wet grinded and polished with silicon carbide paper of 320 to 1200-grit and
ultrasonically degreased with acetone for 15 minutes, rinsed with ethanol, rinsed with distilled
water and dried with pressurized air. The specimen was partially isolated from the environment
with 3M™ polyester tape 8402 with silicone adhesive, leaving a circular area of 2.01 cm2 to react
with the environment. The specimens were finally stored in a desiccator and used in for an
experiment within one hour after storage.
51
Voltamettric tests
Figure 20: Corrosion cell set-up, exploded view with water.
52
Voltamettric tests
5.1.2 Setup
The study was carried out in a three-electrode cell. The reference electrode was a SCE. The
auxiliary electrode was initially a stainless steel mesh but was replaced by graphite rods of at least
two times the surface area of the exposed surface area of the working electrode, as is prescribed
in ASTM G61 [64]. The results presented were attained with the graphite rod as auxiliary
electrode. The working electrode was the specimen as described above.
A variety of standard corrosion cells, to position the electrodes and to set up the polarization test,
are on the market. These are generally rather advanced to enable oxygen free corrosion tests. The
cells that were available in the lab did not facilitate mixing with a stir bar. Therefore, a self-made
corrosion cell was eventually used that consisted of an acrylic container with an electrode holder
mounted on the top to ensure fixed distance between the electrodes. A schematization of the cell
is given in Figure 20 and Appendix E.
Figure 21: Corrosion cell set-up, assembled view without water.
53
Voltamettric tests
5.1.3 Test conditions

The electrolyte was SCW either without inhibitor as base case, or with an inhibitor mixed at a
specific concentration. The same type of inhibitors were used as in described in Section 2.2.4. The
experiments were performed at least three times with the following type of solutions:
Blanc: only SCW;

The prepared specimens were immersed in a constantly stirred volume of 210 mL of one of the
solutions, and hung vertically 1 cm under the free surface of the solution for a total of 6 days at
room temperature. The 6 days of immersion were based on OCP measurements discussed in the
results of the next section. After these 6 days, the OCP of the electrochemical cell was measured
for at least 10 minutes. Consecutively, a CPP test was performed based on ASTM G61[64].
The applied potential of the CPP test started at -20mV vs. OCP, and increased in the more noble
direction with a constant potential scan rate of 1.2 V/h. The scanning direction was reversed at the
preset threshold current of 100 mA. This threshold current is one decade greater than the current
recorded at the breakdown point. When this condition was not met, the scanning direction was
reversed at the preset potential of 1.1 V. The reversed scan was stopped when the original OCP
value was reached again.
The OCP and CPP tests were carried out on potentiostats of multiple brands due to practical
reasons, although a single potentiostat was preferred for repeatability. The first potentiostat was a
Solartron 1287 potentiostat/galvanostat that was connected to Solartron 1281 mulitplexer. The
second one was a Solarton 1286 without a connection to a multiplexer. The third potentiostat was
a Autolab PGSTAT302N and the fourth was a Biologic SP-200.

Post-test visual examinations of the surface were carried out to evaluate the type of corrosion, the
location of the corrosion and the dimensions of the eventual pits. This was done by simple
photographs and by inspection with a Keyence Digital Microscope with a vh-z100r/w/t Wide-
range Zoom lens (100-1000X). The SEM/EDS analysis were carried out using the JSM-IT100
from JEOL at high vacuum and the acceleration of the electron beam was 20.0 kV. The samples
were mounted to the specimens stage with conductive adhesive to prevent the accumulation of
electrostatic charge. The XRD analysis were carried out with a D8 Advance diffractometer from
Bruker using Cu K-α X-ray energy.
54
Voltamettric tests
5.2.1 OCP measurements

The duration of the specimen immersion, before a CPP was carried out, was based on OCP
measurements with the Solartron 1287 in the solutions Blanc, NT0 and Si200. OCP measurements
of eight days indicated a trend that confirmed the expected passivation development. No distinct
difference in the trend was observed between the different solutions. The change rate of the OCP
decreased the most during the first 36 hours. In a few cases, a drop in the OCP was measured
followed by repassivation. This occurred generally during the first two days. Based on the time
for the passivation to develop and the likelihood of a drop in the OCP during the first few days of
immersion, a low OCP change rate was expected at 6 days of immersion. Additional
measurements over a period of 6 days were executed to confirm this. Figure 22 provides an
example of the OCP development measured with the Autolab PGSTAT302N. The trend could be
described by a natural logarithm.
Figure 22: OCP Development of mild steel immersed in SCW.
55
Voltamettric tests
Although the OCP values at the moment of immersion deviated from each other with
approximately 10 mV, the final OCP values after 6 days show a larger deviation. The OCP values
of all the samples that showed passivation behavior, have been summarized in Figure 23. The
average OCP values of all solutions are within each other’s range. So the inhibitors do not have a
significant influence on the final OCP value.
The standards and literature often recommend and use a stabilization period of one hour or a
maximum change rate, such as 3 mV/min as recommended by ASTM F2129 [66]. The specimens
reached this maximum change rate within an hour. However, the specimens would not have had
enough time to form a passivating layer on the surface before the CPP tests. CPP test that would
only test the inhibition efficiency of inhibitors, and not in combination with the natural passivation
of the metal sample, would require one hour of stabilization.
Figure 23: Final OCP values after 6 days of immersion in different solutions.
56
Voltamettric tests
Figure 24: Three types of CPP curves: a) Passive behavior with breakdown of passivity; b) Active behavior followed
by passive behavior and breakdown of passivity; c) Passive behavior without breakdown of passivity.
57
Voltamettric tests
5.2.2 CPP measurements

The CPP measurements that followed the OCP measurements showed three type of polarization
curves, as shown in Figure 24. Figure 24.a shows a polarization curve that is characterized by a
vertical upward branch. Specimens that are sufficiently passivated start in the passivity domain
and show this type of polarization behavior. Figure 24.b shows an steep increase of the current
density at the beginning of the polarization scan, which can be related to specimens that are not
completely passivated and which therefore start in the active domain. The distinct feature of Figure
24.c is the reversed branch that developed on the left side of the upward branch. This behavior is
related to specimens where localized corrosion followed by breakdown of the passivity did not
occur. At high potentials, the oxygen evolution potential can be reached, resulting in the oxidation
of H2O instead of the oxidation of the metal surface and therefore no breakdown of the passivity.
Active behavior, as in Figure 24.c, was often predicted before the CPP test. A small inhomogeneity
at the interface of the polyester tape and the free metal surface that was spotted with close visual
inspection with the naked eye, was an indication for active behavior. This was the case for 2 out
of 27 specimens. CPP tests confirmed these observations. Due to practical problems with the
potentiostats, the results obtained on the solutions NT100, NT50 and Si100 were not representative
and are not further discussed.
Most specimens showed the behavior of Figure 24.a, where the initiation and propagation of
localized corrosion occurred at potentials more electronegative than the oxygen evolution
potential. An overview of the CPP results for the solutions Blanc, NT0 and Si200 is given in Figure
25. All upward anodic polarization scans showed similar slopes and metastability. The Blanc
specimens showed distinct pit initiations and repassivations, which were less distinct for NT0 and
Si200. The reversed branch of three solutions is clearly different, although none exhibited a
protection potential. The Blanc CPP tests showed a jagged branch, which indicates unsuccessful
attempts for repassivation and unstable pits, and followed an almost vertical path without
decreasing in current density. The NT0 and Si200 tests did not have a jagged reverse branch and
did decrease in the current density. The largest decrease in the current density is assigned to the
solution NT0. From these results it was suggested that although the solutions with inhibitors have
a lower breakdown potential, the inhibitors did have a positive effect on the environment at the
pits and prevent pitting propagation.
Lower scan rates resulted in a more vertical upward branch and lower breakdown potentials
because the system had more time to react and to stabilize. Higher scan rates can reduce the
possibility of localized corrosion around gaskets and masks according to Scribner Associates [67].
The metastability can be the result of a too high scan rate. However, it is more likely that the
metastability was the result of insufficient passivation. The insufficient passivation was probably
located at the interface of the polyester tape and the free metal surface. Surface observations
identified crevice corrosion and confirmed the presumed cause of the scattered branch. CPP tests
with the same scan rate, but with specimens were active behavior initially dominated, did not result
in a similar scattered result, as can be seen in Figure 24.
58
Voltamettric tests
Figure 25: Overview of representative CPP curves.
59
Voltamettric tests
The breakdown points of all CPP tests were extracted and summarized in Figure 26a. Some of
these breakdown are not suitable for comparison due to a different scan rate, i.e. the points with
higher breakdown potentials, or due to active behavior, i.e. the points with lower breakdown
potentials. Figure 26b shows the breakdown points that were suitable for comparison, and
demonstrate that the resistance to initiation and propagation of localized corrosion is highest for
the Blanc solution and lowest for the Si200 solution, as in Figure 25. What became clear from
Figure 26b is that a general linear relationship was distinguished of which the equation is given in
the figure. The coefficient of determination (R2) indicates a good fit of the line to the data points.
The effect of the scan rate had most likely a larger influence on the breakdown points than the
corrosion behavior itself. This was pointed out by comparing the slope of the linear fit with the
slope of the anodic branch of the red curve in Figure 25, i.e. b13_Blanc, the values were 0.35
V/μA and 0.36 V/μA, respectively
Change of the pH during the CPP test could have affected the breakdown point. This becomes
more likely when the test volume of the electrolyte is smaller. Measurements of the pH, EC and
temperature before and after the polarization tests were not large enough to affected the breakdown
point.
Figure 26: Overview of the breakdown points of the CPP tests with a) all CPP tests, and b) with only the
representative CPP tests.
60
Voltamettric tests

Post-analysis by visual inspection before and after etching of the specimen surface confirmed that
crevice corrosion took place at the interface of the polyester tape and the free metal surface, as is
shown in Figure 27. 3D-microscopy was used to analyze the shape of the pits that were formed
during the CPP tests. An example of such a 3D image is given in Figure 28. The maximum depth
to average width ratio was calculated and presented for the three solutions in Figure 29. It appears
that the presence of inhibitors resulted in more shallow pits, with the best results for Si200. This
is in agreement with what was previously stated based on shape of the CPP curves, that the
inhibitors reduced the corrosion propagation at the pit location.
Passivation of the metal surface was confirmed not only by OCP and CPP measurements, but also
by XRD and SEM/EDX analysis. SEM/EDX analysis indicated iron oxides for the Blanc and NT0
solutions. Fe2O3 was identified on all samples that were immersed in the Blanc, NT0 and the SiO2
solutions, as is shown in the XRD results in Table 8 and Appendix H. No precipitated silica was
found in the form of quarts. Amorphous silica scales cannot be detected by XRD. Other analysis
techniques are required to assert the presence or absence of silica scale. SEM/EDS is such a
method that should be able to detect silica scale. Some products at the edges of pits after CPP tests
in Si200 solutions indicated the presence of silica, as is shown in Figure 31.b. However, the
adsorbed layers were most likely affected by the removal of the specimens from the solution and
by the introduction of the specimens in high vacuum.
Figure 27: Digital microscopy image at 200x of crevice corrosion at the polyester interface.
61
Voltamettric tests
Figure 28: Digital microscopy images of a pit after a CPP test, a) 3D image, b) top view image with depth profile.
62
Voltamettric tests
Figure 29: The maximum depth to average width ratio of pits after CPP tests.
Table 8: XRD results.
Sample Compound Compound

Blanc Iron Fe
Iron oxide Fe2O3
Iron carbide FeCx
NT0 Iron Fe
Iron oxide Fe2O3
Iron carbide FeCx
Si2000 Iron Fe
Iron oxide Fe2O3
Iron carbide FeCx
63
Voltamettric tests
Figure 30: SEM/EDS results of a) uncorroded surface after CPP in Blanc solution, b) corrosion product outside a pit
after CPP in Si200 solution, c) surface inside a pit after CPP in Si200 solution.
64
6 General discussion and Conclusions
The goal of this thesis was to gain understanding on the corrosion susceptibility of mild steel in
SACW, to test the effect of different corrosion inhibitors on the corrosion susceptibility, and to
explore phosphorus-free corrosion inhibitors for this type of cooling water. This was done through
literature research, brief simulations in PHREEQC and lab experiments. Based on this
information, answers to the research questions raised in Chapter 1.2 are provided in this chapter.
A. How can the corrosion behavior of mild steel in soft alkaline cooling water be evaluated?
First of all, the composition and quality parameters of SACW must be known and, just as
importantly, the water must also be physically available to be used in experiments. Therefore,
SCW was produced of which the composition and quality was based on the data of three surface
water intake points in The Netherlands that are indirectly used for cooling water production. The
concentrations of the ions in the surface water were increased to the desired COC. Simplifications
were made by leaving out trace elements, organic carbon and iron. A natural pH of 9.1 was reached
based on the carbonate equilibrium. The pH was increased by lowering the NaHCO3 dosage and
increasing the Na2CO3 dosage. The Ca2+ and Mg2+ concentrations that are normally present in
surface water were completely exchanged for Na+ for the composition of synthetic soft water, as
if the CIEX had a 100% efficiency in removing the hardness. The calculations and the composition
were validated using CBE calculations, PHREEQC simulations, pH and EC measurements and IC
analysis.
The corrosion assessment techniques started off with desk studies related to corrosion indices and
saturation indices in PHREEQC. Thereafter, mass loss tests and polarization experiments were
carried out. At last, surface analyses were undertaken to verify the results obtained in the other
experiments.
The corrosion indices that appeared to be most suitable for SACW have been evaluated. These
turned out to be inappropriate for this kind of cooling water because they were based too much on
empirical relationships for waters with their characteristics within a specific range.
PHREEQC simulations were used to check the SCW composition, whether it is suitable for the
experiments and which type of scaling products could be expected. The scaling products that
emerge from running the simulation can be evaluated for their benefits and disadvantages for the
protection of the CWS.
A variety of lab experiments are possible to determine the corrosion rate. The most wide applied
tests are mass loss tests. Mass loss tests in SCW can be carried out with different durations and
multiple specimens in a setup of 2L jars to provide an estimate on the corrosion rate and corrosion
type. Determination of the corrosion efficiency based on mass loss is only appropriate for
specimens showing the same type of corrosion. Differentiation should be made between the
inward and the outward facing side of the specimens.
65
General discussion and Conclusions
OCP, PR and CPP tests are also suitable for testing the corrosion of mild steel and SCW is a
suitable medium. These polarization tests can be carried out in simplified corrosion cells with 210
mL of SCW. OCP tests should run for several days if the formation and corrosion resistance of a
passive layer is evaluated. Crevice corrosion can easily distort the measurements of passivated
samples. In case only the performance of inhibitors is evaluated and not on the natural protection
of the metal by an oxide, only one hour of stabilization is required before starting the CPP
measurements. Not only CPP tests, but also Tafel plots can be determined form which the
corrosion rate can be derived and the effect of the inhibitor on the anodic or cathodic curve. This
report focused however on mild steel that was passivated by a naturally formed iron oxide layer.
The creation of Tafel plots requires cathodic polarization which will remove the oxide layer.
Therefore, the corrosion rate was not determined based on Tafel plots.
Visual surface analysis with the naked eye and digital microscopy with a magnification up to
1000x provided information on the surface morphology, i.e. the type of corrosion and the depth-
width ratio of possible pits. The elemental composition found at the surface was identified with
SEM/EDS. XRD is a useful method to identify the crystalline structure of oxide products and
possible precipitates. Possible silica scale products were most likely amorphous, which makes
XRD unsuitable for silica scale identification. The adsorbed layers were most likely affected by
the removal of the specimens from the solution and by the introduction of the specimens in the
high vacuum of the SEM/EDS. More advanced techniques, such as EIS and FTIR have not been
evaluated for this type of water but are likely to provide more insight in the adsorbed layer and the
surface characteristics of the specimen.
B. What is the effect of soft alkaline cooling water on the corrosion behavior of mild steel?
It was derived from the potential-pH diagram for iron in an aerated solution that mild steel would
be naturally passivated by a protective oxide layer at pH 9.3. Two different types of specimens
were used that had slightly different chemical composition but also different roughness. The
specimens with the 1200-grit polish, confirmed the expectations based on the potential-pH
diagram. The specimens with a four times higher roughness showed however uniform corrosion.
The suggested reason for this is the inability of the specimen to form a homogeneous protective
oxide layer on the surface due to the surface roughness.
The 1200-grit polished specimens showed an increase in potential over a period of six days similar
to a natural logarithmic trend. This implicated the development of a passivating layer. Anodic
potentiodynamic polarization, starting at -20mV vs. OCP, reaffirmed this. XRD measurements
also confirmed the presence of protective Fe2O3.
The anodic branch of the CPP tests showed some pit initiations before actual propagation. At a
scan rate of 0.6 V/h the breakdown potential of mild steel in SCW was 0.4 ± 0.02 V vs. SCE. The
reversed branch indicated that the environment at the pits did not prevent further propagation, nor
did the mild steel specimens repassivate in SCW.
66
General discussion and Conclusions
C. How do different non-toxic corrosion inhibitors affect the corrosion behavior of mild steel in
soft alkaline cooling water?
Two type of non-toxic inhibitors were used, namely NovoTraqua® and sodium silicate. The
addition of NT0 and Si200 inhibited uniform corrosion but did not provide complete protection
against localized corrosion in the mass loss tests. Si200 performed certainly the best as inhibitor.
The inhibition efficiency of different inhibitor concentrations has to be tested though, because
localized corrosion can be the result of too low inhibitor concentrations.
The inhibitors did not influence the OCP value after six days of immersion when compared to the
OCP of the Blanc solution. The breakdown potential derived from the CPP tests was highest for
the Blanc solution and lowest for the Si200 solution, which was against expectations. This is most
likely due to a too high scan rate. Lower scan rates resulted in much lower breakdown potentials
for the Blanc solution while the NT0 and Si200 solutions did not have a similar decrease in
breakdown potential.
The CPP tests did not show considerable difference in the anodic upward branch for the specimens
in different solutions. The reversed branch was however less jagged and showed attempt for
repassivation for the inhibited solutions when compared to the Blanc solution. This was related to
the positive effect the inhibitors had on the environment at the pits. The explanation was sought
in the attraction of the negative dipoles of the inhibitors by the positively charged pits providing
protection against aggressive ions. Without inhibitors, only aggressive ions like Cl and SO24
would be attracted by the positive pit, resulting in pit propagation. The pits of the Si200 solution
were most shallow relative to their width, and deepest for the Blanc solution. So the surface
analyses confirmed what was derived from the CPP tests.
D. How do different concentrations of phosphorus in the corrosion inhibitor affect the corrosion
behavior of mild steel in soft alkaline cooling water?
Literature review described that phosphonates catalyze the oxidation of ferrous hydroxide to
gamma iron oxide. Based on this it was likely that the oxide layer at the mild steel surface would
be less protective at lower dosages of phosphonates. So more localized corrosion and lower
breakdown potentials were expected. Tests with different concentrations of phosphonates in the
NovoTraqua® product were not successful due to practical reasons. This research question is
therefore left unanswered and recommendations are provided in the next chapter.
67
7 Recommendations
The results presented in this report may contribute to the decision making and understanding of
other (pilot scale) research projects that seek for alternative and more sustainable operation of
current CWS. The research also resulted in new questions and recommendations for further
research which are discussed below.
The immersion setup should be optimized to allow for automatic temperature regulation of the
water inside the jugs and to allow for automatic inhibitor dosage. This provides the opportunity to
perform corrosion tests at temperatures that generally occur in CWS. It would also provide the
opportunity to test the inhibitor performance at different temperatures. Inhibitors that rely on
physical adsorption are expected to have a lower inhibition efficiency at higher temperatures. So
relative inhibition efficiencies at room temperature may not present the same relative inhibition
efficiencies at higher temperatures.
The last research question was not answered, however it is still worthwhile to research the effect
of different inhibitor concentrations and different phosphorus concentrations within the inhibitors.
This can result in knowledge on the adsorption strength of an inhibitor. The corrosion data would
have to be fitted to various adsorption isotherms, for example the Langmuir and Frumkin
isotherms. The best fit would provide information on the thermodynamics of the inhibitor
adsorption.
Similar tests can be conducted as presented in this report. The dynamic polarization tests would
require only 1 hour stabilization period to test the different inhibitor and phosphorus
concentrations. It is recommended to establish complete Tafel plots to evaluate the effect of the
inhibitor on the cathodic and anodic curve and to determine the corrosion rate. Crevice corrosion
should not influence the measurements. Coating the specimen instead of polyester tape, rubber O-
ring or PTFE mask is recommended.
This study made use of different analytical techniques to identify surface morphology, deposits
and scales. The used techniques were optical microscopy, SEM/EDS, and XRD. More advanced
techniques such as Fourier Transform Infrared Spectroscopy (FTIR) and Electrochemical
Impedance Spectroscopy (EIS) are recommended to obtain better understanding of the adsorbed
layers and the inhibition mechanis. FTIR can be a useful method for obtaining information on the
functional groups, and for the detection of vibrational frequencies of the bonds that are linked to
specific chemical species. It is particularly suitable for the identification of organic substances in
deposits. Additionally, FTIR provides the opportunity to follow surface layer growth. [68] EIS
provides information on the adsorbed layer and the interaction with the metal surface by modelling
the corrosion cell as an electrical circuit with several electrical elements.
69
Recommendations
This research focused on the corrosion behavior of mild steel in SACW. The high pH and
phosphorus limited conditions most likely affect microbiological growth in the CWS. So robust
and reliable operation of a CWS with this type of water does not only require revision of the
corrosion inhibitors, but also revision of the need for dispersants and biocides. This research made
use of SCW with a composition of a COC of 10. It is recommended to run experiments at different
COC to determine the range of the eventual conditioning program and to evaluate of further
reduction of water intake and chemicals can be accomplished.
70
8 Bibliography
[1] D.J. Flynn, The Nalco Water Handbook, third ed., McGraw-Hill Professional, New York,
2009.
[2] P.A. Saini, I.H. Farooqi, A. Hussain, M.A. Quraishi, Study of low cost eco‐friendly
compounds as corrosion inhibitors for cooling systems, Anti-Corrosion Methods Mater. 46
(1999) 328–335. doi:10.1108/00035599910295508.
[3] I.T. Miettinen, T. Vartiainen, P.J. Martikainen, Phosphorus and bacterial growth in drinking
water., Appl. Environ. Microbiol. 63 (1997).
[4] BiOPhree, (n.d.). http://www.aquacare.nl/nl/ (accessed July 6, 2018).
[5] F. Oesterholt, H. Huiting, N. Koeman-Stein, L. Paping, Cooling water conditioning in the
future, 2016. https://www.tkiwatertechnologie.nl/wp-content/uploads/2015/08/KWR-
2016.069-Cooling-Water-Conditioning-in-the-Future-V3.0_final_public.pdf (accessed
March 23, 2017).
[6] P.R. Roberge, P.E.N. York, C. San, F. Lisbon, L. Madrid, M. City, M.N. Delhi, S. Juan,
Corrosion Engineering : Principles and Practice, (2008). doi:10.1036/0071482431.
[7] J.E. Singley, The search for a corrosion index, J. Am. Water Works Assoc. (1981) 579–582.
[8] W.F. Langelier, THE ANALYTICAL CONTROL OF ANTI-CORROSION WATER
TREATMENT, J. Am. Water Works Assoc. 28 (1936) 1500–1521.
http://www.jstor.org/stable/41226418.
[9] R.V. Skold, T.E. Larson, Measurement of the instantaneous corrosion rate by means of
polarization data, Corrosion. 13 (1957) 69–72.
[10] T.M. Riddick, The mechanism of corrosion of water pipes, Water Sew. Work. 91 (1944).
[11] R. Touir, M. Cenoui, M. El Bakri, M. Ebn Touhami, Sodium gluconate as corrosion and
scale inhibitor of ordinary steel in simulated cooling water, Corros. Sci. 50 (2008) 1530–
1537. doi:10.1016/J.CORSCI.2008.02.011.
[12] B. Zhang, C. He, X. Chen, Z. Tian, F. Li, The synergistic effect of polyamidoamine
dendrimers and sodium silicate on the corrosion of carbon steel in soft water, Corros. Sci.
90 (2015) 585–596. doi:10.1016/J.CORSCI.2014.10.054.
[13] J. Marín-Cruz, R. Cabrera-Sierra, M.A. Pech-Canul, I. González, EIS study on corrosion
and scale processes and their inhibition in cooling system media, Electrochim. Acta. 51
(2006) 1847–1854. doi:10.1016/J.ELECTACTA.2005.02.104.
[14] D.J. Choi, S.J. You, J.G. Kim, Development of an environmentally safe corrosion, scale,
and microorganism inhibitor for open recirculating cooling systems, Mater. Sci. Eng. A.
335 (2002) 228–235. doi:10.1016/S0921-5093(01)01928-1.
[15] M. Saremi, C. Dehghanian, M.M. Sabet, The effect of molybdate concentration and
hydrodynamic effect on mild steel corrosion inhibition in simulated cooling water, Corros.
Sci. 48 (2006) 1404–1412. doi:10.1016/j.corsci.2005.06.009
71
Bibliography
[16] Y.W. Kim, J.G. Kim, D.J. Choi, Development of a blended corrosion, scale, and
microorganism inhibitor for open recirculating cooling systems, Mater. Corros. 52 (2001)
697–704.
[17] M.A. Ahmed, Inhibition of mild steel corrosion in cooling systems by low-and non-toxic
corrosion inhibitors, The University of Manchester, 2016.
https://www.research.manchester.ac.uk/portal/files/60828940/FULL_TEXT.PDF
(accessed April 6, 2018).
[18] J.P. Broomfield, Corrosion of steel in concrete: understanding, investigation and repair,
CRC Press, 2003.
[19] M. Finšgar, J. Jackson, Application of corrosion inhibitors for steels in acidic media for the
oil and gas industry: A review, Corros. Sci. 86 (2014) 17–41.
doi:10.1016/j.corsci.2014.04.044.
[20] S. Nešic, N. Nešic´, Key issues related to modelling of internal corrosion of oil and gas
pipelines-A review, (2007). doi:10.1016/j.corsci.2007.06.006.
[21] M. Pourbaix, Atlas of electrochemical equilibria in aqueous solutions, Pergamon Press,
New York, 1966.
[22] M. Poubaix, Potential-pH diagrams and metallic corrosion, in: W.H. Ailor (Ed.), Handb.
Corros. Test. Eval., John Wiley & Sons, New York, 1971: pp. 661–687.
[23] U.R. Evans, Evidence in favor of an electrochemical mechanism of corrosion, J. Inst. Met.
30 (1923) 265–267.
[24] U.R. Evans, Isolation of the Film Responsible for the Passivity of an Iron Anode in Acid
Solution, Nature. 126 (1930) 130.
[25] J.E.O. Mayne, M.J. Pryor, The Mechanism of Inhibition of Corrosion of Iron by Chromic
Acid and Potassium Chromate., J. Chem. Soc. (1949) 1831.
[26] J.E.O. Mayne, J.W. Menter, M.J. Pryor, The Mechanism of Inhibition of the Corrosion of
Iron by Sodium Hydroxide Solution, J. Chem. Soc. (1950) 3229.
[27] M. Cohen, The passivity and breakdown of passivity on iron, Passiv. Met. Ed. by RP
Frankenthal, J. Kurger, NJ Princieton, Electrochem. Soc. (1978) 521.
[28] H.-H. Strehblow, Passivity of Metals Studied by Surface Analytical Methods, a Review,
Electrochim. Acta. 212 (2016) 630–648. doi:10.1016/j.electacta.2016.06.170.
[29] S. Haupt, H.-H. Strehblow, Corrosion, Layer Formation, and Oxide Reduction of Passive
Iron in Alkaline Solution: A Combined Electrochemical and Surface Analytical Study
(1986) 873–885.
[30] ASTM International, G046 - Standard Guide for Examination and Evaluation of Pitting
Corrosion, ASTM Int. (2013). doi:10.1520/G0046-94R13.
[31] D.A. Jones, Principles and prevention of corrosion, 2nd ed., Prentice-Hall, 1996.
72
Bibliography
[32] R.A. Pisigan Jr, J.E. Singley, Influence of buffer capacity, chlorine residual, and flow rate
on corrosion of mild steel and copper, J. Am. Water Works Assoc. (1987) 62–70.
[33] T.E. Larson, R.M. King, Corrosion by water at low flow velocity, Corrosion. 10 (1954)
110–114.
[34] T.E. Larson, R.V. Skold, Laboratory studies relating mineral quality of water to corrosion
of steel and cast iron, Corrosion. 14 (1958) 43–46.
[35] T. Fujii, T. Kodama, H. Baba, Influence of Water Quality and Flow Conditions on
Corrosion of Carbon Steel Pipes in Fresh Water, Corros. Eng. 31 (1982) 637–642.
[36] I. Matsushima, Carbon steel corrosion in freshwaters, in: R.W. Revie (Ed.), Uhlig’s Corros.
Handb., Third, John Wiley & Sons, Inc, 2011: pp. 589–607.
[37] V.S. Sastri, Corrosion inhibitors: principles and applications, John Wiley & Sons, 1998.
[38] H.H. Uhlig, Uhlig’s corrosion handbook, Third, John Wiley & Sons, Hoboken, New Jersey,
2011.
[39] M. Prabakaran, K. Vadivu, S. Ramesh, V. Periasamy, Corrosion protection of mild steel by
a new phosphonate inhibitor system in aqueous solution, Egypt. J. Pet. 23 (2014) 367–377.
doi:10.1016/J.EJPE.2014.09.004.
[40] B. Zhang, C. He, X. Chen, Z. Tian, F. Li, The synergistic effect of polyamidoamine
dendrimers and sodium silicate on the corrosion of carbon steel in soft water, Corros. Sci.
90 (2015) 585–596. doi:10.1016/j.corsci.2014.10.054.
[41] O. Lahodny-Šarc, L. Kaštelan, The influence of pH on the inhibition of corrosion of iron
and mild steel by sodium silicate, Corros. Sci. 21 (1981) 265–271. doi:10.1016/0010-
938X(81)90002-0.
[42] L. Lehrman, H.L. Shuldener, Action of Sodium Silicate as a Corrosion Inhibitor in Water
Piping, Ind. Eng. Chem. 44 (1952) 1765–1769. doi:10.1021/ie50512a023.
[43] M. Stern, A.L. Geary, Electrochemical polarization I. A theoretical analysis of the shape of
polarization curves, J. Electrochem. Soc. 104 (1957) 56–63.
[44] ASTM International, G031 - Standard guide for laboratory immersion corrosion testing of
metals, ASTM Int. (2012) 1–10. doi:10.1520/G0031-12A.
[45] ASTM International, G003 - Conventions Applicable to Electrochemical Measurements in
Corrosion Testing, ASTM Int. G3-14 (2014) 1–10. doi:10.1520/G0003-14.2.
[46] NEN, NVN6419:2008: Protocol for the preparation of synthetic laboratory samples for
organic compounds in ground- and surface water, (2008).
[47] United Nations, Globally harmonized system of classification and labelling of chemicals
(GHS), New York and Geneva, 2011.
https://www.unece.org/fileadmin/DAM/trans/danger/publi/ghs/ghs_rev04/English/ST-SG-
AC10-30-Rev4e.pdf (accessed June 18, 2017).
73
Bibliography
[48] ASTM International, D1141-98(2013) Standard Practice for the Preparation of Substitute
Ocean Water 1, (2013). doi:10.1520/D1141-98R13.
[49] R.L. Malcolm, P. MacCarthy, Limitations in the use of commercial humic acids in water
and soil research, Environ. Sci. Technol. 20 (1986) 904–911.
[50] R.L. Malcolm, Geochemistry of stream fulvic and humic substances, Humic Subst. Soil,
Sediment, Water Geochemistry, Isol. Charact. (1985) 181–209.
[51] G.B. Alexander, W.M. Heston, R.K. Iler, THE SOLUBILITY OF AMORPHOUS SILICA
IN WATER, (1954). https://pubs.acs.org/doi/pdf/10.1021/j150516a002 (accessed June 1,
2017).
[52] C.A.J. Appelo, SPECIFIC CONDUCTANCE: how to calculate, to use, and the pitfalls,
(2010). http://www.hydrochemistry.eu/exmpls/sc.html (accessed August 19, 2017).
[53] P. Atkins, J. De Paula, Atkins’ physical chemistry, New York. (2006) 77.
[54] David L. Parkhurst, C.A.J. Appelo, Description of Input and Examples for PHREEQC
Version 3—A Computer Program for Speciation, Batch-Reaction, One-Dimensional
Transport, and Inverse Geochemical Calculations, 2013.
ftp://brrftp.cr.usgs.gov/pub/charlton/phreeqc/Phreeqc_3_2013_manual.pdf (accessed
August 19, 2018).
[55] I. Safari, M. Hsieh, S. Chien, M.E. Walker, R.D. Vidic, D.A. Dzombak, J. Abbasian, Effect
of CO2 stripping on pH in open‐recirculating cooling water systems, Environ. Prog.
Sustain. Energy. 33 (2014) 275–282.
[56] H.S. Harned, R. Davis Jr, The ionization constant of carbonic acid in water and the solubility
of carbon dioxide in water and aqueous salt solutions from 0 to 50, J. Am. Chem. Soc. 65
(1943) 2030–2037.
[57] H.S. Harned, S.R. Scholes Jr, The ionization constant of HCO3-from 0 to 50, J. Am. Chem.
Soc. 63 (1941) 1706–1709.
[58] D.L. Gallup, P. von Hirtz, Control of Silica-Based Scales in Cooling and Geothermal
Systems, in: Miner. Scales Depos. - Sci. Technol. Approaches, 2015: pp. 573–582.
[59] J.R. Davis, Corrosion: Understanding the basics, ASM International, 2000.
[60] ASTM International, G001 - Standard Practice for Preparing , Cleaning , and Evaluating
Corrosion Test, ASTM Int. 90 (2011) 1–9. doi:10.1520/G0001-03R11.2.
[61] ASTM International, G005 - Standard Reference Test Method for Making Potentiodynamic
Anodic Polarization Measurements, ASTM Int. (2014) 1–9. doi:10.1520/G0005-14.2.
[62] 3M België N.V. & 3M Nederland B.V., Productcatalogus tapes - tape oplossingen voor de
industrie, (2010) 164.
http://solutions.3mnederland.nl/3MContentRetrievalAPI/BlobServlet?lmd=138493876200
0&locale=nl_NL&assetType=MMM_Image&assetId=1319209837063&blobAttribute=I
mageFile (accessed May 6, 2018).
74
Bibliography
[63] ASTM International, D2688-Standard Test Method for Corrosivity of Water in the Absence
of Heat Transfer (Weight Loss Method) 1, (2015). doi:10.1520/D2688-15E01.
[64] ASTM International, G061 - Standard Test Method for Conducting Cyclic Potentiodynamic
Polarization Measurements for Localized Corrosion Susceptibility of Iron-, Nickel-, or
Cobalt-Based Alloys 1, (2014). doi:10.1520/G0061-86R14.
[65] ASTM International, F746 Standard Test Method for Pitting or Crevice Corrosion of
Metallic Surgical Implant Materials, (2014). doi:10.1520/F0746-04R14.
[66] ASTM F2129, F2129 Standard Test Method for Conducting Cyclic Potentiodynamic
Polarization Measurements to Determine the Corrosion Susceptibility of Small Implant
Devices, (2017). doi:10.1520/F2129-17B.
[67] Scribner, CView Manual, (1996).
[68] C.P. East, T.L. Schiller, C.M. Fellows, W.O.S. Doherty, Analytical Techniques to
Characterize Scales and Deposits, in: Miner. Scales Depos. Sci. Technol. Approaches,
2015: pp. 681–700.
75
9 Abbreviations
AE auxiliary electrode
ASTM American Society for Testing and Materials
CBE charge balance error
CIEX cation exchange
COC cycles of concentration
CPP cyclic potentiodynamic anodic polarization
CWS cooling water system
EC electrical conductivity
EDS energy dispersive spectroscopy
EIS electrochemical impedance spectroscopy
FTIR Fourier transform infrared spectroscopy
HTI holding time index
IEX ion exchange
KWR KWR Watercycle Research Institute
NT NovoTraqua®
OCP open circuit potential
ORCWS open recirculating cooling water system
77
Appendix A – HumVi production and composition
Vitens N.V. is a drinking water company that uses groundwater to produce drinking water by the
following consecutive treatment processes: vacuum degassing, plate aeration, rapid sand filtration,
CO2 removal, pellet softening, caustic soda dosage, another rapid sand filtration, and finally ion
exchange for color removal. The ion exchange tank is regenerated with a 10% NaCl solution. The
depleted regenerate stream contains ca. 4% salt solution and a high concentration of TOC. To
reach a product with a high TOC concentration and low salt concentration, the regenerate stream
is concentrated by dead-end nanofiltration, nanofiltration and diafiltration. The composition of the
final product HumVi is summed up in Table 9. (P. Sjoerdsma, personal communication, August
30, 2017)
Table 9: Composition of HumVi, a product with concentrated humic substances (Vitens N.V.).
(HS=humic substances, i.e. humic and fulvic acids; See end of document for other abbreviations.)
Percentage
Unit Value Unit Value
of TOC
Colour [mg PtCo/L] 456691 TOC [mg/L] 7.56 100%
TOC [mg/L] 101514 DOC [mg/L] 6.92 91,5%
pH 8.33 POC [mg/L] 0.63 8,3%
EC [mS/cm] 36.7 HOC [mg/L] 0.29 3,8%
dry matter [g/L] 228
Salt/NaCl [%] 0.85 HS [mg/L] 2.92 38,6%
UV [1/cm] 353103 BB [mg/L] 2.32 30,7%
NO3- [mg/L] 450
HCO3- [mg/L] 38402 LMW acids [mg/L] 7.56 100%
CO32- [mg/L] 1172 LMW neutr [mg/L] 6.92 91,5%
PO43- [mg/L] 17
SO42- [mg/L]
Cl- [mg/L] 5150
Si [mg/L] 200
Na+ [mg/L] 34991
K+ [mg/L] 14.2
Fe [mg/L] 121
Mn [mg/L] 4.58
Ca2+ [mg/L] 57
Mg2+ [mg/L] 59
79
Appendix B – PHREEQC input file
Database C:\phreeqc\database\phreeqc.dat
#Defining Solution 1
SOLUTION 1 Simulated hard surface water starting at COC1
units mg/l
pH 8 CO2(g) -3.1
temp 20
Alkalinity 160 as HCO3
Na 32
K 8.0
#Ca 60 as Ca
Ca 150 as CaCO3
#Mg 8 as Mg
Mg 35 as CaCO3
Amm .1
Fe(+3) .1
Mn 0.01
S(6) 52 #SO4-2
N(+5) 13 #NO3-
P .4 as PO4
Cl 46
F .2
Br .09
Si 2 as SiO2
END
#Defining Solution 2
SOLUTION 2 Simulated soft surface water starting at COC1
units mg/l
pH 8 CO2(g) -3.1
temp 20
Alkalinity 160 as HCO3
Na 116
K 8.0
Ca 0
Mg 0
Amm .1
Fe(+3) .1
Mn 0
Cu(+2) 0
S(6) 52.0 #SO4-2
N(+5) 13.0 #NO3-
P .4 as PO4
Cl 46.0
F .2
Br .09
Si 2 as SiO2
END
#Increasing the COC of Solution 1 at T=20oC

USE solution 1
REACTION_TEMPERATURE 20
81
Appendix B – PHREEQC input file
EQUILIBRIUM_PHASES
CO2(g) -3.1
REACTION 1
H2O -1
52.733 moles in 53 steps #COC20
USER_GRAPH 1
-headings CoC SI_Calcite SI_Dolomite SI_Sepiolite SI_Talc pH
-axis_titles "COC" "SI" "pH"
-initial_solutions false
-chart_title "Impact of COC on SIs and pH of hard water at 20oC"
-plot_concentration_vs x
-axis_scale x_axis 1 20 2
-axis_scale y_axis -5 15 5
-axis_scale sy_axis 8 9.5 0.5
-start
10 graph_x 1/SOLN_VOL
20 graph_y si("Calcite")
30 graph_y si("Dolomite")
40 graph_y si("Sepiolite")
50 graph_y si("Talc")
60 graph_sy -la("H+")
-end
END
USER_GRAPH 1
-detach
END
#Increasing the COC of Solution 2 at T=20oC

USE solution 2
REACTION_TEMPERATURE
20
EQUILIBRIUM_PHASES
CO2(g) -3.1
REACTION 1
H2O -1
52.733 moles in 53 steps #COC20
USER_GRAPH 2
-headings CoC SI_Chalcedony SI_Quartz pH
-axis_titles "COC" "SI" "pH"
-initial_solutions false
-chart_title "Impact of COC on SIs and pH of soft water at 20oC"
-plot_concentration_vs x
-axis_scale x_axis 1 20 2
-axis_scale y_axis -1 1 0.2
-axis_scale sy_axis 8 9.5 0.5
-start
10 graph_x 1/SOLN_VOL
20 graph_y si("Chalcedony")
30 graph_y si("Quartz")
40 graph_sy -la("H+")
-end
END
USER_GRAPH 2
-detach
(B.M van Breukelen, personal communication, October 09, 2017)
82
Appendix C – 4 weeks immersion test setup
Figure 31: Schematization of the 4 weeks immersion tests setup.
83
Appendix D – Specimen composition and surface roughness
Table 10: Chemical composition of the specimens used in the Immersion tests and the Voltamettric tests.
Weight percent in Weight percent of material
Compound
Immersion tests in Voltamettric tests
C 0.08 0.087
M 0.330 0.375
P 0.011 0.013
S 0.009 0.013
Al 0.045 NA
Si 0.019 0.018
Fe remainder remainder
Figure 32: Surface roughness of Metal Samples Co coupon vs. self-made coupon.
85
Appendix E – Corrosion cell setup
Figure 33: Line drawing of self-made corrosion cell
87
Appendix F – CPP test results
Figure 34: Overview of the CCP tests per solution type, with a) Blanc, b) NT0, and c) Si200.
89
Figure 35: Overview of the CCP tests with the top two rows Blanc, third row NT0, and fourth row Si200.
90
Table 11: Overview of breakpoint or vertex indices, current densities and potentials.
Name Index Current Potential

Density (V vs. SCE)
(μA/cm2)
a1_Blanc 918 0.680 0.666
a2_Blanc 1218 1.114 0.952
a3_NT0 992 0.734 0.746
a4_NT0 522 0.265 0.332
a5_Si200 614 0.474 0.372
a6_Si200 1122 0.848 0.823
b10_Blanc 3708 1.808 0.842
b11_NT0 4419 2.479 0.962
b12_Si200 1422 0.111 0.222
b13_Blanc 3054 1.631 0.709
b14_NT0 1685 0.357 0.316
b15_Si200 1043 0.057 0.141
b16_Blanc 963 0.043 0.204
b17_NT0 1427 0.193 0.255
b18_Si200 2025 2.136 0.166
c2_Blanc 3207 0.628 0.390
c3_Blanc 3582 0.944 0.419
91
Appendix G – Surface visualization after CPP tests
Figure 36: Overview of the surfaces after CCP tests per solution type, with column a) Blanc, column b) NT0, and
column c) Si200.
93
Appendix H – XRD results
Figure 37: XRD measurements of mild steel coupons after immersion in a) Blanc, b) NT0 and c) Si200 solutions.
95

Tesis Corrosion, Codigos Phreeqc

Uploaded by

Copyright:

Available Formats

Tesis Corrosion, Codigos Phreeqc

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tesis Corrosion, Codigos Phreeqc

Uploaded by

Copyright:

Available Formats

Preface

Soli Deo gloria.

cycles of concentration. Better understanding of the protection mechanisms would be obtained

Appendix A – HumVi production and composition..................................................................... 79

1.1 Background and Problem description

Figure 1: Tetrahedron of undesired interacting processes in ORCWS.

1.2 Objective and Research Questions

D. How do different concentrations of phosphorus in the corrosion inhibitor affect the

1.3 Research approach and Report Outline

2.1 Open recirculating cooling systems

Figure 2: Schematization of an Open Recirculating Cooling Water System.

Figure 3: Makeup rate and blowdown as a function of COC.

2.2 Corrosion in aqueous environments

2.2.1 Thermodynamics and kinetics of iron corrosion

The anodic oxidation reaction of a metal in general:

O2  2H 2O  4e  4OH  E0 = + 0.40 V SHE (2.6)

Figure 4: Schematic corrosion cell of iron in an aqueous environment.

The electrochemical processes can be displayed in a thermodynamic framework based on

2H   2e  H 2 E0 = 0.0 V SHE (2.7)

O2  4H   4e  2H 2O E0 = + 1.23 V SHE (2.8)

Fe  Fe2  2e E0 = -0.44 V SHE (2.9)

K s  (aFe2  )( aOH  ) 2  1.64  10 14 at 18°C (2.11)

2Fe II (OH )2  2OH   Fe III 2O3  H 2O  2H 2O  2e (2.12)

2.2.2 Localized corrosion

2.2.3 Water parameters affecting corrosion

2.2.4 Corrosion inhibitors

Another way of categorizing corrosion inhibitors is by subdividing the inhibitors in three

Figure 7: Classification of inhibitors (based on diagram in Uhlig’s Corrosion Handbook [38])

2.3 Corrosion measurements

2.3.1 Corrosion indices

The index can be interpreted by the guidelines given in Table 1.

Table 1: Interpretation guidelines for the LSkI

CO2 Free CO2 Concentration (mg/L)

Table 2: Interpretation guidelines for the RI

2.3.2 Mass loss measurements

K = a constant for calculating the corrosion rate in the right units;

Figure 8: Hypothetical cathodic and anodic polarization diagram (ASTM G3 [45]).

2.3.4 Surface analysis

3.1 Materials and Methods

3.1.1 Synthetic cooling water production

Table 4: Chemicals used to create the stock solutions to produce SCW.

Three stock solutions were proposed:

Stock A: Bulk solution with soluble salts;

3.2 Results and Discussion

3.2.1 SCW composition

3.2.3 Corrosion indices

4.1 Materials and Methods

4.1.1 Coupon preparation and composition

Figure 14: Photo of the 4 weeks immersion test setup.

4.1.3 Test conditions

Blanc: only SCW;

Inhibitor concentration development based on HTI

Startup dosage (ul)

Figure 17: Inhibitor concentration development due to inflow of SCW and

4.1.4 Surface analyses

Table 7: Rating table for pit density.

1. Coupon preparation as in Section 4.1.1;

At the startup of the 4 weeks immersion test:

During the 4 weeks:

At the end of the 4 weeks immersion test: