Offshore prosessanlegg og rørledninger-materialtekniske utfordringer Norsk Forening for vedlikehold, Bergen 16-17 oktober 2007

Hvordan behandle CO2/H2S i design av olje- og gassystemer

Arne Dugstad, Institute for Energy Technology

N-2027 Kjeller, Norway Fax: +47 63 80 62 58

E-mail: arne.dugstad@ife.no Carbon steel is thermodynamically unstable in water with dissolved CO2 and the only reason that carbon steel is so attractive and can be so widely used in oil and gas production is that the steel surface becomes covered by a protective layer of corrosion products, oil, mineral scale or inhibitors. It is relatively easy to predict and explain the high corrosion rates on bare steel. The real challenge is to reduce the corrosion and that requires knowledge about the performance of the protective layers, means to predict the breakdown of the layers and methods and techniques to ensure that robust layers form on the surface. Appendix A discusses how CO2 affects the water chemistry, the electrochemical reactions on the bare steel surface, and the initiation and growth of protective corrosion product films. As many sweet systems contain organic acids that affect the solution chemistry and the formation and stability of the FeCO3 corrosion product films, organic acids need also to be considered when the effect of CO2 is discussed. As discussed in Appendix A corrosion of carbon steel is complex and involves a large number of different corrosion rate controlling mechanisms. These mechanisms are controlled by different parameters and respond very differently to changes in the environment. Several prediction models for corrosion of oil and gas pipelines have been developed. Some of the models are based on mechanistic modelling of the different processes involved, while other models are mainly based on empirical correlations with laboratory or field data. Common for all models is that they require a large number of more or less well defined input parameters and that they miss some parameters that are obviously important for the corrosion process. Appendix B gives a brief overview of most of the available prediction models and discusses the consequences of lack of reliable input parameters and negligence of important corrosion controlling parameters and processes.

- 1 - Appendix A

Appendix A: Fundamental Aspects of CO2 Metal Loss Corrosion

Arne Dugstad, Institute for Energy Technology

INTRODUCTION

The mechanism of carbon steel corrosion in a CO2 containing environment has been studied and debated for decades. Hundreds of papers related to CO2 corrosion have been published and a large variety of corrosion rates and mechanisms have been reported. Oil companies and research institutions have analyzed the data and developed a number of prediction models1 to take account of the various parameters that determine the corrosion rate. The models give up to two decades difference in the predicted CO2 corrosion rate and it all depends on how the various parameters are treated and how much conservatism that is built into the model.

In order to explain the confusion and the apparently contradictory observations and results that have been seen and reported, it is important to realize that the term “CO2 corrosion” and the effect of CO2 is not related to one mechanism only. A large number of CO2 dependent chemical, electrochemical and mass transport processes occur simultaneously on and close to the corroding steel surface. The various reactions respond differently to changes in CO2 partial pressure, temperature, water chemistry, flow and other operational parameters. All the reactions should be taken into account when corrosion in a CO2 containing environment is to be quantified and explained.

Many researchers have studied and discussed the electrochemical reactions taking place on the bare steel surface. The mechanisms that control the rate of the electrochemical reactions are of great academic interest, but are less important when it comes to the practical application of carbon steel. When carbon steel is directly exposed to water and CO2 the bare steel corrosion rate will under almost all circumstances become prohibitively high for practical use in oil and gas production. This is illustrated in Figure 1 where the corrosion rate has been predicted2 for various CO2 partial pressures and pH values as a function of temperature. The corrosion rate predicted up to 40 °C apply for bare steel, while partly protective films are formed at higher temperature. It is seen that the corrosion rates are in the order of several mm/year, even at CO2 partial pressures below 0.5 bar, i.e. pressures where the old “rule of thumb” says that carbon steel can be applied without any treatment3.

In the present paper it is focused on fundamental corrosion mechanisms in sweet systems. Three major effects of CO2 will be addressed: The effect on the water chemistry, the effect on the electrochemical reactions, and the impact on the initiation and growth of corrosion product films. As many sweet systems contain organic acids that affect the solution chemistry and the formation and stability of the FeCO3 corrosion product films, organic acids also have to be included in the discussion. The effect of other parameters is discussed in the paper “Fundamental Aspects of CO2 Metal Loss Corrosion, Part II: Influence of different Parameters on the CO2 Corrosion Mechanism”4

- 2 – Appendix A

CO2 AND THE EFFECT ON WATER CHEMISTRY AND pH

When CO2 is dissolved in water it is partly hydrated and forms carbonic acid:

3222 COHOHCO ⇔+ (1)

Carbonic acid is diprotic and dissociates in two steps:

H CO H HCO2 3 3⇔ ++ − (2)

−+− +⇔ 233 COHHCO (3)

The resulting pH is a function of the CO2 partial pressure. This is illustrated in Figure 2 where the pH has been calculated # as a function of CO2 partial pressure in unbuffered water and in water with 1, 10 and 100 mM alkalinity respectively. The pH decreases with increasing CO2 partial pressure. When the water is buffered, the pH increases but the dependency of the CO2 partial pressure follows the same trend as for pure water.

When the steel corrodes, Fe2+ and an equivalent amount of alkalinity are released in the corrosion process.

232

32 22 HHCOFeCOHFe ++→+ −+ (4)

The pH in the solution increases and when the concentrations of Fe2+ and CO32- ions exceed the

solubility limit, precipitation of FeCO3 can occur:

)(323

2 sFeCOCOFe ⇒+ −+

(5)

When solid FeCO3 is formed at the same rate as the steel corrodes, the pH becomes constant in the corroding system.

The solubility of FeCO3 is strongly dependent on the pH and the CO2 partial pressure. This is illustrated in Figure 3 where the amount of Fe2+ needed to be produced by corrosion to reach FeCO3 saturation is plotted as function of pH at the start of the corrosion process (1 wt% NaCl). It is seen that [Fe2+]sat is much higher in condensed water than in typical formation water with a pH above 5. When the system is pH stabilised at pH 6.5-7.5, [Fe2+]sat is only a fraction of a ppm and is reduced 100 times per unit pH increase. It is important to note that the solubility curves in Figure 3 are not parallel and they cross in the range pH 4.5 to 5. The consequences are illustrated in Figure 4 where the pH is kept constant while the Fe2+ solubility is plotted as a function of CO2 partial pressure. It is seen that the solubility of Fe2+ goes through a maximum for pH 5 and 5.5. At lower pH the solubility increases with increasing CO2 partial pressure while the trend is opposite at pH 6. As the pH is often used as an

# Sol1.81: IFE in-house pH and solubility program developed for water and mixed water and glycol systems

- 3 – Appendix A

indicator for the corrosivity of produced water it should be noted that corrosion films form more easily at pH 5 with very high CO2 partial pressures (>> 1 bar) than with 1 bar CO2

5. When comparing the corrosivity of various waters, a higher pH will always give less contribution from the H+ reduction reaction, but the pH cannot be used directly to predict the likelihood for formation of protective corrosion product films.

The presence of HAc and other organic acids makes the water chemistry much more complex. The interaction of HAc has been discussed in detail in two recent review papers by Crolet and Bonis6 and by Gulbrandsen7. HAc is a weak acid that is readily soluble in water. The equilibrium HAc partial pressure over a 1 mM HAc solution is less than 1 Pa at temperatures up to 100 °C. As pointed out by several authors7,8, HAc is often referred to as a stronger acid than carbonic acid in the literature. The source of this confusion is that dissolved CO2, and not only H2CO3, is often erroneously included in the term “carbonic acid”. The dissociation constant Ka of H2CO3 according to eq. (2), expressed as pKa, is about 3.5 at 25 oC, i.e. lower than for HAc, which has a pKa of about 4.8. Accordingly, H2CO3, which is the main cathodic reactant in CO2 corrosion,9-11 is a stronger acid than HAc. A consequence of this is that H2CO3 has a higher reaction rate constant than HAc8.

The concentrations of HAc in the water can typically be a few mM. Crolet and Bonis12 have proposed a classification where less than 0.1 mM HAc is regarded to give only slightly increased risk of corrosion, while concentrations of more than 1 mM significantly increases the corrosivity. Undissociated HAc concentrations of more than 10 mM have been reported in some fields.

Acetic acid dissociates according to equation 6:

−+ +⇔ AcHHAc (6)

The three equilibria in eqs. 2,3 and 6 are linked through the same H+ concentration and form a two- buffer system. The shift in pH due to the presence of acetic acid and the effect on FeCO3 solubility can be calculated by solving these equations together with eq. 5 for FeCO3 solubility. The shift in [Fe2+]sat at 0.1, 1 and 10 mM HAc respectively is given in Figure 5. The relative increase in solubility is highest at low pH. The increase was about 0.2, 3 and 8 times when the undissociated HAc concentration was 0.1, 1 and 10 mM respectively in condensed water.

As discussed by Gulbrandsen7, Fe2+ and Ca2+ can form acetate complexes13,15 according to eqs. 7-9:

Fe2+ + Ac- = FeAc+ (7)

Fe2+ + 2Ac- = FeAc2 (8)

Ca2+ + Ac– = CaAc+ (9)

In formation waters with high content of calcium, a significant fraction of Ac- can therefore be present as CaAc+. Since the concentration of bicarbonate and HAc+Ac- is used to estimate the in situ pH in

- 4 – Appendix A

formation water, it might be important to take complex formation into account, particularly for high Ca2+ brines and high temperatures.

ELECTROCHEMICAL REACTIONS AT THE BARE STEEL SURFACE

Cathodic reactions

A large number of papers have proposed and discussed various cathodic and anodic reactions on the steel surface in the presence of CO2. Most of these reactions have recently been reviewed, discussed and summarized by Nesic et al11,14,16 and Garsany et al8. The three main cathodic reactions appear to be:

221 HeH →+ −+ (10)

−− +→+ 3221

32 HCOHeCOH (11)

−− +→+ AcHeHAc 221 (12)

Equations 10-12 are the overall reaction routes and do not indicate the detailed mechanisms of the proton reduction. The relative contribution from each of these reactions depends on the concentrations, temperature, pH, convection etc. The presence of CO2 affects all the reactions directly or indirectly by affecting the H+ concentration and the amount of undissociated HAc and H2CO3.

In strong acids, which are fully dissociated, the rate of hydrogen evolution occurs according to eq 10 and cannot exceed the rate at which H+ ions are transported to the surface from the bulk solution (mass transfer limit). Its contribution to the corrosion rate is small above pH 5, a typical pH in formation water. H2CO3 serves as an additional source of H+ ions17 that enables the hydrogen evolution reaction to proceed at a much higher rate than in a solution of a strong acid at the same pH.

Although it is still debated, it seems to be widely accepted that cathodic reduction of protons from molecular H2CO3, also referred to as “direct H2CO3 reduction”, may take place8,11,18-22 (eq 11) and thus contribute to an increase in the corrosion rate beyond the limits of the H+ reduction rate. One suggested mechanism is that H2CO3 adsorbs and reacts on the electrode18,19,20. Another proposed mechanism is formation of H2 from direct reduction of H2CO3. The rate determing step is assumed to be slow hydration of CO2

8,11,21-24, at least at temperatures below 50-60°C. It was found to be activation controlled above this temperature11. It should be noticed that the term “direct reduction” has not been well defined and the actual mechanism is unclear.

It has been well documented that the presence of HAc has a detrimental effect on the corrosion rate and the morphology of the attack in sweet systems when the concentration of undissociated HAc exceeds 0.1-1 mM. The effect of organic acids was debated in the early 40’ies, and then almost forgotten before the topic was revitalized by Crolet and Bonis in the 80’ies. Recently Crolet and Bonis presented a review paper that sums up much of their work6. In this paper it is focused on the effect of the water chemistry, the pH and how the stability of the iron carbonate films is reduced in the presence of HAc. Crolet and co-workers found that HAc increased the cathodic limiting current25 and that the anodic reaction was inhibited by HAc26. Although HAc affects both the cathodic and anodic

- 5 – Appendix A

reaction mechanisms, the effects have been regarded as secondary, with minor consequences for the corrosion attack seen in the field6.

Over the last 10 years there has been a growing awareness amongst operators and researchers for the accelerated corrosion caused by organic acids. A large number of papers have addressed the HAc issue and tried to shed more light on the mechanisms. Most of these papers have been discussed and summarized in Gulbrandsen’s review7.

Little work addresses directly the basic effects of HAc on the anodic and cathodic reactions. The increased cathodic limiting current and the inhibition of the anodic reaction reported by Crolet et al have been seen also by other workers7,8,9,10. Gulbrandsen7 reported that localised attacks were formed at low temperature. This was attributed to the inhibition of the anodic reaction, not the formation of a corrosion product film.

As for the case of H2CO3, it has not been fully agreed whether HAc is directly reduced at the surface or only acts as an extra source of H+. Sun et al reported that the H2 evolution from HAc was activation controlled at room temperature and that HAc acts solely as an additional source of hydrogen ions. Garsany et al did voltammetry studies that showed two waves related to H+ and HAc reduction respectively8,9. Since the dissociation of HAc is very fast, it was not possible to distinguish between the reduction of HAc and the reduction of H+ after dissociation.

It has been suggested that direct reduction of the bicarbonate ion27,28 and water11,29 can become important at low pCO2 partial pressure and high pH:

−−− +→+ 2323 222 COHeHCO (13)

−− +→+ OHHeOH 222 22 (14)

The concentration of HCO3- increases with pH and becomes more than 300 times the H2CO3

concentration at 1 bar CO2 and pH 6. Although the concentrations can be high, the low dissociation constant makes bicarbonate (pK ca. 10) and water poor proton donors and it is believed that the contribution from these reactions will be negligible under normal sweet conditions. However it is difficult to experimentally distinguish the effect of this particular reaction mechanism for hydrogen evolution from eqs 10 and 11.

A high bicarbonate concentration might also affect the corrosion rate by enabling the regeneration of the reactant (H2CO3)30. It is therefore not necessary to consider bicarbonate reduction to explain the observed increase in current at pH>6.

−− +⇔ 233232 COCOHHCO (15)

Anodic reactions

The overall anodic dissolution of iron is given in equation 16:

−+ +→ eFeFe 22 (16)

- 6 – Appendix A

This reaction, which has a pH dependent rate in accordance with Bockris experimental results33, has been extensively studied and the various reaction steps have been discussed in detail by Drazic31 and Lorenz and Heusler32. The mechanism suggested by Bockris et al.33 for strong acids

−+ ++⇔+ eHFeOHOHFe 2 (17a)

−+ +⇔⎯→⎯ eFeOHFeOH rds (17b)

OHFeHFeOH 22 +⇔+ +++ (17c)

has frequently been assumed to apply in CO2 solutions18,20,21,28. As pointed out by Nesic14, it has been overlooked that the experimental data indicated that the pH dependency decreased rapidly with increasing pH. The reaction order with respect to OH- was 2 at low pH but decreased towards 1 and 0 at pH>4. In a study by Nesic et al.34 it has been confirmed that the anodic dissolution of iron does not depend significantly on OH- concentration above pH 4, but is affected by the presence of CO2, as previously indicated by Davies and Burstein35 and Videm36.

The following iron dissolution rate equation was derived from experimental results and proposed by Nesic et al34:

[ ] ( ) abaCO

aa pOHki 102

2

1−=E

(18)

where for:

pH<4 → a1 = 2 ba=0.03 V per decade

4<pH<5 → a1 = 2-0 ba=0.03-0.12 V per decade

pH>5 → a1 = 0 ba=0.12 V per decade

pCO2 < 10-2 bar → a2 = 0

10-2 <pCO2 < 1 bar → a2 = 1

pCO2 > 1 bar → a2 = 0

They also proposed a mechanism to explain the experimental results at pH>5:

LFeCOFe ⇔+ 2 (19a) −+ ++⇔+ eHOHFeOHFe adLL 2 (19b)

- 7 – Appendix A

−+ +⎯→⎯ eOHFeOHFe adLrds

adL (19c) ++ +⇔+ HOHFeOHOHFe adLadL 22 )( (19d)

solLadL OHFeOHFe 22 )()( ⇔ (19e)

OHCOFeHOHFe solL 222

2 22)( ++⇔+ ++ (19f)

where FeL denotes the complex Fe-CO2. They assumed that the complex is formed as an adsorbed species at the electrode surface that catalyses the dissolution of iron. For pH<4, it is postulated that the rate of charge transfer step 19c will increase, and the desorption step 19e will then be rate-determining. The problem in such model consideration is to document the existence of the inter mediates.

CO2 AND HOW IT AFFECTS THE PROTECTIVE FILM FORMATION

Film formation is a complex process and the precipitation rate of iron carbonate is the main controlling factor in sweet systems. The morphology and the composition of the film determine whether an attack develops as worst case corrosion, low corrosion with protective films or mesa corrosion. The layer also influences the corrosion inhibitor access and availability at the surface and therefore plays an important role for inhibitor performance37,38

A number of papers address the film properties and how the films are formed. Much of this work has recently been discussed and referenced by Kermani39. In the present paper it will only be focused on some fundamental principles and trends.

The driving force for precipitation is the supersaturation of FeCO3. The precipitation rate is generally slow, and a high degree of supersaturation of iron carbonate in the water is necessary in order to get sufficient amounts of iron carbonate deposits on the steel surface. Whether it deposits or not depends amongst other factors on the kinetics of precipitation and on how it is anchored to the surface.

In principle there are two steps involved in the precipitation processes; nucleation and particle growth. It is assumed that the rates of these processes are related to the relative supersaturation (RS) where S is the supersaturation, Q is the concentration of the solute at any instant, Qeq the equilibrium solubility and Ksp the solubility product for FeCO3

sp

COFe

K

CCS

−+ ⋅=

23

2

(20)

)1( −=−

= SQ

QQRS

eq

eq (21)

- 8 – Appendix A

The rate of nucleation is believed to increase exponentially with relative supersaturation, whereas the rate of particle growth bears an approximately linear relationship to this parameter. Growth should therefore predominate at low relative supersaturation. When the supersaturation is high, the exponential dependency of nucleation rate may cause this process to occur to the near exclusion of particle growth and a colloidal solution might form close to the steel surface or inside the film. Coagulation might therefore be important for protective film formation under these conditions. Coagulation is accelerated by heating, and increasing ionic strength.

Calculated supersaturation for a sweet system with various CO2 partial pressures is given as a function of the pH in the brine in Figure 7. It is assumed that a steady state corrosion process is established that gives 30 ppm dissolved Fe2+in the water, and that the pH referred to is the pH in the brine before corrosion products are dissolved. It is seen that the supersaturation in a brine with a certain pH increases substantially when the CO2 partial pressure is increased. The higher supersaturation gives a higher precipitation rate and increased likelihood for film initiation and formation. The results fit with the observation that a higher pH is required for protective film initiation/formation when the CO2 partial pressure is low.

Fe2+ will be produced in the solution at the metal surface while the H+ concentration is reduced on the steel surface or on the film surface if the film is of the conducting type (iron carbide). The resulting concentration gradients might give a higher supersaturation in the film and on the metal surface and that will increase the precipitation rate locally. These effects have been modeled in detail by Nesic et al14,16 and have been included as an integral part of the corrosion models developed at IFE16 and Ohio University40.

It has been reported several studies on the growth rate of FeCO341,42,43. Calculated growth rates

based on Johnson and Tomson’s work41 is shown as an example in Figure 6. It is seen that the predicted growth rate is very low at 20 °C, but increases fast when temperature and supersaturation increase.

The general description of the growth and nucleation mechanisms given above fit with many of the observations in the authors laboratory44. Figure 8 shows some corrosion films formed in various experiments. At high temperature (> 60 °C) the precipitation rate is fast and the supersaturation low. Under these conditions dense crystalline films are formed which often give good protection. At lower temperature (< 40 °C ) a much lower precipitation rate is experienced, and the relative supersaturation can therefore become very high when dissolved iron carbonate is accumulated. Under these conditions where the relative supersaturation is high a corrosion film with low crystallinity is expected. The expectation is consistent with the appearance of the films actually formed, which are porous, loosely adherent, more or less non crystalline and much less protective than those formed at higher temperatures.

The critical part of the film forming process is the initiation. If the temperature and the supersaturation in the bulk phase are high, a massive precipitation takes place and a dense iron carbonate film can form on most steel surfaces. At lower temperatures and precipitation rates,

- 9 – Appendix A

however, protective film may never form if the steel surface is cleaned (sand blasted, descaled) and continuously exposed. A stagnant period or a semi dry period where high local supersaturation can be achieved44 may therefore be necessary in order to start the deposition of iron carbonate on the steel surface. Experiments have indicated that quenched and tempered low carbon steel are more dependent on such a trigger mechanism than ferritic pearlitic steels, where the pearlite colonies can form a porous network on the surface suitable for anchoring iron carbonate precipitate when the steel is corroded. Once a protective film has formed, however, there is apparently no systematic difference in the ranking of the two steel types when it comes to the development of localized attack. Since small differences in the steel composition and the microstructure can affect the anchoring properties and the development of localized attack, it is not surprising that large differences in the performance of similar carbon steels have been seen in both field applications and laboratory work.

Protective film formation is accelerated by all measures which restricts the transport of reaction products from the surface and which can anchor the corrosion product. Freshly ground specimens studied in the laboratory are therefore much more susceptible to high corrosion rates compared to real pipelines, where an oxide layer is grown during rolling and storage. It can be discussed whether it is the oxide films, which keep the corrosion rate low in the long run, or if the oxide films play an intermittent role only as an initiator for protective iron carbonate films.

Experiments have shown that an apparently dense corrosion film is attached directly to the metal surface when good protection is obtained, while a porous film, sometimes filled with iron carbonate in the outer part only, is formed when poor protection is obtained. A mechanism for how iron carbonate film can form close to the metal surface has been suggested42. It is assumed that the growth rate of iron carbonate has to be equal or higher than the corrosion rate. If the growth rate is less than the corrosion rate, the gap between the iron carbonate filled part of the film and the steel surface will never be filled with dense iron carbonate, and the corrosion film thus formed will be porous or non adherent, see Figure 8b. In order to get an iron carbonate growth rate matching a corrosion rate of 1 mm/y at 40 °C, about 50 times supersaturation is necessary according to Figure 6. Such high supersaturation can apparently be established close to the steel surface when the dissolved iron content is high and a surface layer like rust, mill scale etc restrict the transport of reactants and corrosion product. The maximum supersaturation will then be a balance between the corrosion rate, nucleation rate and the rate of transport.

Once a reasonably dense corrosion product film has formed it can be speculated on whether it is the reduced transport through the iron carbonate film or the formation of a passive film which reduces the corrosion rate45. Experiments showed that the steel shows passive-like behavior at high pH. High pH can be obtained close to the steel when the transport of reactants and corrosion products are restricted by the deposited corrosion products.

CONCLUSIONS

Carbon steel is unstable in water with dissolved CO2 and the only reason that carbon steel can be so widely used is that the surface becomes covered by a protective layer of oil, corrosion products, mineral scale or inhibitors.

- 10 – Appendix A

CO2 is assumed to contribute to the cathodic reaction rate in two different ways. H2CO3 can be directly reduced and dissociation of H2CO3 can serve as a source of H+ ions. The rate determing step is the CO2 hydration reaction below 50-60 °C. Activation control has been reported at higher temperatures.

The presence of CO2 does not seem to affect the anodic reaction significantly in the pH range 4-6 when the acetate concentration is low.

Acetic acid serves as an additional source of H+. It is uncertain whether it also can be directly reduced.

The presence of CO2 affects all the proposed catodic reactions directly or indirectly by affecting the H+ concentration and the amount of undissociated HAc and H2CO3. When a sweet system contains HAc, the relative contribution from the various reactions depends on the concentrations, temperature, pH, convection etc.

The solubility and precipitation of FeCO3 is a function of a large number of parameters where pCO2, HAc+Ac- concentration, temperature and the amount of dissolved corrosion products are the most important parameters. A computer program is necessary in order to account for all the parameters. Some general trends and special effects are:

• The amount of FeCO3 that can be dissolved in a brine goes through a maximum at around 1-2 bar CO2 for a given pH in the range 4.5-5.5.

• The supersaturation of FeCO3 at a given pH and Fe2+ concentration increases with increasing CO2 partial pressure.

• Formation of the CaAc+ complex can give a too low estimation of pH in high Ca-brine if it is not compensated for the Ac-bounded to the complex.

Increased understanding of the CO2 material loss mechanisms can increase the application range of carbon steel and reduce the risk for loss of the pipeline due to misjudgment and wrong treatment. Research areas that should be addressed in the future includes:

• Fundamental studies of film growth in complex water chemistry

• Interaction between corrosion films and oil. How are the growth and the morphology of the corrosion film affected? How is oil wetting affected by the various solid products that can accumulate on the steel surface?

• Inhibitor performance on steel covered or partly covered with corrosion products, scale and other deposits.

- 11 – Appendix A

ACKNOWLEDGMENT

I wish to thank IFE (Institute for Energy Technology) for permission to publish the paper and my colleges Egil Gulbrandsen, Jon Kvarekvål and Rolf Nyborg for helpful suggests in the preparation of the paper.

REFERENCES

1 Rolf Nyborg: Overview of CO2 Corrosion Models for Wells and Pipelines. CORROSION/2002, Paper No. 02233, (Houston, TX: NACE International, 1993).

2. "CO2 Corrosion Rate Calculation Model, Rev. 2", NORSOK standard No. M-506, http://www.standard.no/petroleum, (Oslo: Standards Norway, 2005).

3 “Corrosion in Oil and Gas-Well Equipment” (Dallas, TX;API, 1958)

4. G. Schmitt, M. Hörstemeier, “Fundamental Aspects of CO2 Metal Loss Corrosion, Part II: Influence of different Parameters on the CO2 Corrosion Mechanism” CORROSION/2006, Paper no. 112, (Houston, TX: NACE International, 2006).

5 K. Videm, A. Dugstad, “Film Covered Corrosion, film Breakdown and Pitting Attack of Carbon Steels in aquas CO2 Environments” , CORROSION/88, Paper no. 186, (Houston, TX: NACE International, 1988).

6 J. L. Crolet, M.R. Bonis, “Why so Low Free Acetic Acid Treshold in Sweet Corrosion at Low pCO2“, CORROSION/2005, Paper no. 272, (Houston, TX: NACE International, 2005).

7 E. Gulbrandsen, K. Bilkova, “ Solution Chemistry effects on Corrosion of Carbon Steels in Presence of CO2 and Acetic Acid”, CORROSION/2006, Paper no. 364, Houston, TX: NACE International, 2006).

8 Y. Garsany, D. Pletcher, D. Sidorin, W.M. Hedges, Corrosion, 60, 1155 (2004).

9 Y. Garsany, D. Pletcher, B. Hedges, “The Role of Acetate in CO2 Corrosion of Carbon Steel: Studies Related to Oilfield Conditions”, CORROSION/2003, Paper no. 324, Houston, TX: NACE International, 2003).

10. Y. Sun, K. George, S. Nesic, “The Effect of Acetic Acid on Localized CO2 Corrosion in Wet Gas Flow”, CORROSION/2003, Paper no. 327, Houston, TX: NACE International, 2003).

11 S. Nesic, J. Postlethwaite, S. Olsen: “An electrochemical model for prediction of corrosion of mild steel in aqueous carbon dioxide solutions”, Corrosion, 52 (1996), p 280.

12 R. Bonis, J.-L. Crolet, "Basics of the Prediction of the Risks of CO2 Corrosion in Oil and Gas Wells", CORROSION/89, paper no. 466, (Houston Texas: NACE International, 1989).

13 D.A. Palmer, K.E. Hyde, “An experimental determination of ferrous chloride and acetate complexation in aqueous solutions to 300 °C ”, Geochimica et Cosmochimica Acta, 57, 1393 (1993).

14. S.Nesic, “A Critical Review of CO2 Corrosion Modelling in the Oil and GAS Industry”, 10th Middle East Corrosion Conference 7-10 March 2004.

- 12 – Appendix A

15 J.S.Seewald, W.E.Seyfird, “Experimental determination of portlandite solubility in H2O and acetate solutions at 100-350 °C and 500 bars”, Geochimica et Cosmochimica Acta, 55, 659 (1991).

16. S. Nesic, J. M. Nordsveen, R. Nyborg, A. Stangeland, “A Mechanistic Model for CO2 Corrosion with Protective Iron Carbonate Films”, CORROSION/2001, Paper no. 1040, (Houston, TX: NACE International, 2001).

17. W. Schwenk, Werkstoffe und Korrosion. 25 (1974): p. 643.

18. G. Schmitt, B. Rottman, Werkstoffe und Korrosion. 28 (1977): p. 816.

19. W. Fisher, W. Siedlarek. Werkstoffe und Korrosion. 28 (1978): p. 822.

20. C. de Waard and D. E. Milliams, Corrosion, 31 (1975): p.131.

21. L. G. S. Gray, B. G. Anderson, M. J. Danysh, P. G. Tremaine, “Mechanisms of Carbon Steel Corrosion in Brines Containing Dissolved Carbon Dioxide”, CORROSION/89, paper no. 464, (Houston, TX: NACE International, 1989).

22. E. Eriksrud, T. Søntvedt, "Effect of Flow on CO2 Corrosion Rates in Real and Synthetic Formation Waters", Advances in CO2 Corrosion, Vol. 1. Proceedings of the CORROSION/83 Symposium on CO2 Corrosion in the Oil and Gas Industry, Editors: R. H. Hausler, H. P. Goddard, p. 20, NACE, 1984.

23. S. Nesic, B.F.M. Pots, J. Postlethwaite, N. Thevenot, The Journal of Corrosion Science and Engineering, 1, (1995), paper no. 3. http://www.jcse.org/Volume1/paper3/v1p3.html

24 Ikeda et al, Adv in CO2 corrosion Vol.1 , NACE 1984.

25 J. L. Crolet, M.R. Bonis, “The Role of Acetate in CO2 Corrosion”, CORROSION/83, Paper no. 160, (Houston, TX: NACE International, 1983).

26 J. L. Crolet, N. Thevenot, A. Dugstad, “Role of Free Acetic Acid on the CO2 Corrosion of Steels”, CORROSION/99, Paper no. 24, (Houston, TX: NACE International, 1999).

27. G.I Ogundele, W.E. White, Corrosion 43 (1987): p. 665.

28. L. G. S. Gray, B. G. Anderson, M. J. Danysh and P. R. Tremaine, “Effect of pH and Temperature on the Mechanism of Carbon Steel Corrosion by Aqueous Carbon Dioxide”, CORROSION/90, paper no. 40, (Houston, TX: NACE International, 1990).

29. P. Delahay, J. Am. Chem. Soc., 74, (1952): p. 3497.

30. S. Turgoose, R. A. Cottis and K. Lawson “Modelling of Electrode Processes and Surface Chemistry in Carbon Dioxide Containing Solutions” Computer Modelling In Corrosion, Raymon S. Munn, ASTM Publishing Code Number 04-011540-27.

31. D. M. Drazic, “Iron and its Electrochemistry in an Active State”, Aspects of Electrochemistry, Vol 19, p.79, Plenum Press, 1989.

32. W. Lorenz and K Heusler, “Anodic Dissolution of Iron Group Metals”, in Corrosion Mechanisms, ed. F. Mansfeld (Marcel Dekker, New York, 1987).

33. J. O. M. Bockris, D. Drazic and A. R. Despic, Electrochimica Acta, 4 (1961): p.325.

- 13 – Appendix A

34. S. Nesic, N. Thevenot, J.-L. Crolet, D. Drazic, "Electrochemical Properties of Iron Dissolution in the Presence of CO2 - Basics Revisited", CORROSION/96, paper no. 3, (Houston, TX: NACE International, 1996).

35. H. Davies and G. T. Burstein, Corrosion, 36 (1980) p. 385.

36. K. Videm, "Fundamental Studies aimed at Improving Models for Prediction of CO2 Corrosion", Progress in the Understanding and Prevention of Corrosion, Proceedings from 10th European Corrosion Congress, Vol. 1, p.513 (Institute of Metals, London, 1993).

37. E.Gulbrandsen, S.Nesic, A.Stangeland, T.Burchardt, B.Sundfær, S.M.Hesjevik, S.Skjerve, “Effect of Precorrosion on the Performance of Inhibitors for CO2 Corrosion of Carbon Steel”, CORROSION 98, paper no. 13, (Houston, TX: NACE International, 1998).

38. M.Foss, K.Bilkova, E.Gulbrandsen, M.Knag, J.Sjöblom, “Interaction of CO2 corrosion inhibitors with corrosion product deposits”, Submitted and accepted for presentation at 10th European Symposium on Corrosion and Scale Inhibitors, Ferrara, Italy (2005).

39. M. B. Kermani, A.Morshed, “Carbon Dioxide Corrosion in Oil and Gas Production- A Compendium”, Corrosion, Vol. 59 (2003) p. 659.

40. S. Nesic, J. Cai, K. L. J. Lee, "A Multiphase Flow and Internal Corrosion Prediction Model for Mild Steel Pipelines", CORROSION/2005, Paper No. 05556, (Houston, TX: NACE International, 2005).

41. M.L. Johnson and M.B. Tomson, "Ferous Carbonate Precipitation Kinetics and Its Impact on CO2 Corrosion", CORROSION/91, paper no. 268, (Houston, TX: NACE International, 1991)..

42. E.W.J. van Hunnik, B.F.M. Pots and E.L.J.A. Hendriksen, "The Formation of Protective FeCO3 Corrosion Product Layers in CO2 Corrosion". CORROSION/96, paper no. 6, (Houston, TX: NACE International, 1996).

43. K. Chokshi, W. Sun, S. Nesic, COROSSION/2005, Paper no. 285, (Houston, TX: NACE International, 2005)

44. A. Dugstad, "Mechanism of Protective Film Formation During CO2 Corrosion of Carbon Steel", CORROSION/98, paper no. 31, (Houston, TX: NACE International, 1998).

45. J. L. Crolet, N. Thevenot, S. Nesic, "Role of Conductive Corrosion Products on the Protectiveness of Corrosion Layers", CORROSION/96, Paper No. 4, (Houston, TX: NACE, 1996).

- 14 – Appendix A

0.1

1

10

100

0 20 40 60 80 100Temperature

Cor

rosi

on ra

te /

(mm

/yea

r)

5bar, pH=4 5bar, pH=5.5 2bar, pH=4 2bar, pH=5.50.5bar, pH=4 0.5bar, pH=5.5

FIGURE 1. Calculated2 corrosion rates as a function of temperature at 0.5, 2 and 5 bar CO2 partial pressure and pH 4 and 5.5 respectively. Wall shear stress 20 Pa.

- 15 – Appendix A

3

3.5

4

4.5

5

5.5

6

6.5

7

7.5

8

0 2 4 6 8 10 12 14pCO2 /bar

pH

0 mM1 mM10 mM100 mM

FIGURE 2. pH as a function of CO2 partial pressure in water with various concentration of

NaHCO3. Temperature 60 °C, NaCl =1 wt%

0.01

0.1

1

10

100

1000

3 4 5 6 7 8pH

Fe2+

/ pp

m

0.5 bar1 bar5 bar10 bar20 bar

FIGURE 3. Amount of Fe2+ needed to be produced by corrosion to reach FeCO3 saturation,

plotted as function of pH in the brine. The brine pH is the pH before corrosion has started. (1 wt% NaCl).

- 16 – Appendix A

0

10

20

30

40

50

60

70

80

90

0 1 2 3 4 5 6pCO2 / bar

Fe2+

/ pp

m pH 4.5pH 5pH5.5pH6

FIGURE 4. Amount of Fe2+ needed to be produced by corrosion to reach FeCO3 saturation, plotted as function of pCO2 for various pH’s in the brine. The brine pH is the pH before corrosion has started. (3.5 wt % NaCl, 60 °C).

0.1

1

10

100

1000

3.0 4.0 5.0 6.0 7.0 8.0pH

Fe2+

/ pp

m 1 bar1 bar, 0.1 mM1 bar, 1 mM1 bar, 10 mM

FIGURE 5. Amount of Fe2+ needed to be produced by corrosion to reach FeCO3 saturation, plotted as function of pH in the brine. The given brine pH is the pH before corrosion has started. 1 bar CO2 and 0, 0.1 1 and 10 mM HAc respectively. 1 wt% NaCl,

- 17 – Appendix A

0.01

0.1

1

10

100

1000

20 30 40 50 60 70 80 90

Temperature C

Gro

wth

rate

mm

/yea

r

10X20X50X100X

FIGURE 6. Calculated41 growth rate of iron carbonate as a function of temperature and different supersaturation. NaCl = 1 wt%.

0.1

1

10

100

1000

4 4.5 5 5.5 6 6.5 7pH

Supe

rsat

urat

ion

(S)

0.5 bar1 bar2 bar5 bar10 bar

FIGURE 7. Supersaturation obtained in the brine when it is assumed that the corrosion process releases 30 ppm dissolved Fe2+ in the water. The pH is the pH in the brine before corrosion starts. NaCl=1 wt%.

- 18 – Appendix A

a

b

c FIGURE 8. Corrosion films formed under various conditions. a) 40 °C and SR<40 in bulk

solution, b) 40 °C and SR > 40 in bulk solution, c) 80 °C and SR < 10 in bulk solution.

1 Appendix 2

Corrosion Prediction Tools for Oil and Gas Pipelines - Reliability, Limitations & Challenges

A. Dugstad, J. Kvarekvål and R. Nyborg, Institute for Energy Technology

1.0 INTRODUCTION

When carbon steel is directly exposed to water and even moderate amounts (< 0.5 bar) of CO2 or H2S it becomes unstable and will dissolve at a rate of many mm/year. The reason that carbon steel can be used at all in oil and gas production is that corrosion is effectively reduced or eliminated when the steel becomes covered by a layer of oil, corrosion product and/or inhibitor films. The challenge in corrosion rate prediction is to predict when the various protective layers will form and to predict the performance and the stability of the layers. Corrosion of carbon steel is influenced by a large number of parameters. Some of the most important are temperature, CO2 and H2S partial pressure, flow regime and velocity, pH of the water phase, concentration of dissolved corrosion products, presence of acetic acid, water wetting, and the microstructure, composition and surface condition of the steel. The importance of the various parameters differs depending on the dominating corrosion mechanism and on the objective of the prediction. If the objective is to predict worst case corrosion (corrosion without protective layers) the rate controlling parameters are different from those that control the formation of protective corrosion products films, inhibition and oil wetting. The detailed influence of many of the parameters is still not very well understood and some of them are closely linked to each other. A small change in one of them may influence the corrosion rate considerably. For instance, if the amount of corrosion product in the water phase is varied while temperature (90 °C), CO2 partial pressure (2 bar), velocity (4 m/s) and pH (5.0) are kept constant, corrosion rates in the range <1 mm/year to 20 mm/year can be obtained. This is attributed to changes in the properties of the thin layer of corrosion products and scale that always accumulates on the steel surface, see Figure 2. The morphology and composition of this layer determine whether the attack is worst case corrosion, low corrosion with protective films or mesa corrosion with localized attack. The layer also interacts with inhibitor transport to the surface and therefore plays an important role for inhibitor performance. The oil companies and different research institutions have developed a large number of CO2 corrosion prediction models. Very different results can be obtained when the models are run for the same cases due to the different philosophies used in the development of the models. Some of the models predict corrosion rates based on full water wetting and little protection from corrosion product films. These models have a built-in conservatism and they probably over predict the corrosion attack significantly for many cases. Other models are partly based on field data and predict generally much lower corrosion

2 Appendix 2

rates. In these models it is assumed that reduced water wetting and/or formation of protective scale can reduce the corrosion rate from many mm/year to less than 0.1 mm/year. Some of the models are based on mechanistic modelling of the different chemical, electrochemical and transport processes involved. Other models are mainly based on empirical correlations with laboratory or field data. However, the mechanistic models are usually tuned against lab data to some degree, while the laboratory and field data models often have some mechanistic equations as a starting point. A list of the most used models is given in Table 1 together with an indication of the application limits for the different models. All these models are basically CO2 corrosion models. Several of the models take into account the effect of H2S or organic acid on the pH calculation, but most of the models are not intended for use in situations where H2S or organic acids dominate the corrosion process. Due to large uncertainties in the model predictions and lack of understanding of the corrosion process, carbon steel is often disregarded under conditions where it could be adequate. Better understanding and better prediction tools can therefore increase the application range of carbon steel and have a large economic impact. A joint industry project where the various CO2 corrosion prediction models in Table 1 were compared has been conducted at Institute for Energy Technology 1 .In this project field data with actual corrosion measurements were gathered from the participating oil companies. The corrosion models were run for the field cases, and predicted corrosion rates were compared with the actual measured corrosion rates. The project is continuing by developing a corrosion field database where corrosion field data from several oil companies and several oil and gas fields around the world are collected. A brief discussion of the models that has been evaluated is given in the paper. In order to understand the differences between the models and the large differences in the predicted corrosion rates, it is important to have an understanding of what type of corrosion processes the models are expected to predict. The main group of processes are discussed in the paper.

2. WHAT ARE THE CORROSION PROCESSES WE ARE SUPPOSED TO PREDICT 2.1 "Worst Case" CO2 Corrosion In this paper, worst case corrosion is defined as the corrosion when no protective layers are formed. When a bare steel surface is exposed in a defined environment a fixed corrosion rate could be expected, but this is not the case as the steel surface can accumulate some products that accelerate corrosion up to 3-4 times. The porous layer shown in Figure 1a is iron carbide (Fe3C) from the steel that remains un-attacked on the steel surface when the iron in the steel corrodes and dissolves. This type of films, which does not give any protection, can under certain conditions increase the corrosion rate due to galvanic effects. When very small amounts of corrosion products deposit on the surface, or a porous film like the one shown in Figure 1a is formed, generally very high uniform corrosion rates are obtained, even at low temperatures and high pH. The general trend for the effect of CO2 content, temperature, flow and pH under such conditions is illustrated in Figure 2. This "worst case" corrosion is the easiest type to model and reproduce in the laboratory. Several prediction models have been developed for this type of corrosion and the models predict the corrosion rate usually within a factor of 2.

3 Appendix 2

2.2 Corrosion with Protective Iron Carbonate Films When the solubility product for iron carbonate and other scales is exceeded these constituents can precipitate and deposit on the steel surface giving a dense and protective film as indicated in Figure 1b. It is seen that the porous film shown in Figure 1a has been filled with FeCO3 precipitate close to the metal. The prediction of corrosion rates when films with protective properties are formed is much more complicated than worst case predictions due to the stochastic nature of film formation. The challenge then is to predict if film formation is kinetically hindered, if the film can be removed mechanically, and if new film will form when a protective film has been destroyed locally. The film formation is strongly dependent on the solubility and precipitation of iron carbonate, which again is influenced by temperature, pH, flow rate and the composition and microstructure of the corroding steel. The effect of steel composition and micro structure has generally been given little attention and is not usually included to any extent in prediction models, see sections 5.2 and 5.3. The effect of solubility, precipitation and supersaturation of iron carbonate has been studied in detail at IFE, and the most important findings have been published elsewhere2. Experiments have shown that precipitation of FeCO3 is a slow, temperature dependent process, and that a high degree of supersaturation can be maintained in a corroding system. The precipitation of FeCO3 is facilitated by increased pH, increased temperature, reduced flow velocity and by measures that reduce the transport of reactants and corrosion products to and from the steel surface. 2.3 Localized Corrosion in CO2 Environment Localized corrosion or mesa corrosion is the most feared type of corrosion attack in practice. This attack is characterized by the formation of severely corroded regions separated with sharp steps from neighbouring areas with much less attack, as illustrated in Figure 33. Mesa attack develops when the protective corrosion film on the steel surface is destroyed locally (Figure 1c and Figure 3). The attack is usually associated with high flow rates. However, experiments have shown that localized attack can also be initiated and propagate under semi-stagnant conditions4. The mechanism is based on a fine balance between film growth and corrosion with subsequent film removal. The corrosion rates in the mesa attacked areas are often somewhat higher than those obtained on steel without protective films. The higher corrosion rate is attributed to the galvanic cell established between the non-attacked and the attacked areas. The actual penetration rate of localized corrosion is a useful parameter to know, but for practical purposes it is more important to understand the underlying mechanism and to be able to predict when and where localized corrosion will be initiated, and how it can be prevented. This is one of the important challenges for future mechanism studies and modelling. 2.4 Corrosion in CO2 - H2S Mixtures Extensive research has been conducted all over the world on corrosion and stress corrosion cracking in H2S containing environments. Small amounts of H2S in an otherwise CO2 dominated system effect the corrosion rate because iron sulfides precipitate and form corrosion product films much easier than iron carbonate. A large variety of iron sulfide films can form. This is illustrated in Figure 4 which shows the major iron sulfide inter-relationships in aqueous solutions and their biogenic counterparts5. Mackinawite, pyrrhotite and pyrite are iron sulfides frequently encountered in oil- and gas-related sour corrosion. Mackinawite is a thermodynamically metastable form of iron sulfide, but the formation

4 Appendix 2

kinetics are rapid and mackinawite is generally believed to form as the initial sour corrosion product. A solid state type of reaction between the H2S in solution and the iron rather than as a precipitation reaction of ferrous sulfide from the solution has been proposed as the formation mechanism6. In non-buffered systems (gas condensate pipelines) with moderate and low H2S pressures a thin layer of mackinawite would be expected to form. It has been proposed that the corrosion rate at 1 bar H2S and room temperature is controlled by the combined FeS dissolution rate and corrosion product diffusion through the film7. On the other hand, for very low H2S partial pressures (0.001 bar) it has been reported that mass transport of reacting species (H+, H2S and H2CO3) towards the steel/film surface is the corrosion rate controlling mechanism8. This was explained by the formation of a thin conductive FeS layer on the steel surface, which depolarised the cathodic reactions. At higher concentrations of H2S and HS- ions thermodynamically stable iron sulfides, pyrrhotite and pyrite, is often detected as corrosion products. These sulfides may give very protective films and low corrosion rates. Localised corrosion attacks frequently occur during sour corrosion, and SEM examination of locally corroded areas, as well as electrochemical studies, have shown that corrosion can be sustained beneath thick, but porous FeS deposits. A mechanism has been proposed9 where the external FeS surface is the cathode and where the anodic reaction beneath the FeS deposit is facilitated by a thin layer of aqueous FeCl2 at the Fe/FeS interface. This intervening FeCl2 layer prevents precipitation of FeS directly on the corroding steel surface and thereby enables the anodic reaction to be sustained. In practice, traces of polysulfide species, which are readily electro-reduced, are thought to be present in the H2S/NaCl solutions and to contribute to initiation and propagation of localised corrosion. The CO2 concentration has little effect on the corrosion rate and the corrosion kinetics as long as the H2S concentration is large enough to maintain a stable iron sulfide scale. When a local breakdown of the film occurs the maximum local corrosion rate seems to be proportional to the CO2 partial pressure. Furthermore, it has been reported that maximum corrosion rate is estimated to be approximately 0.5 times the de Waard CO2 corrosion rate10. In CO2-H2S-systems that are buffered or contain large amounts of dissolved corrosion products, competitive precipitation between iron carbonate and iron sulfide can be seen when the H2S concentration is low. The transition from CO2 dominated to H2S dominated corrosion is complex and both temperature and pH dependent. Increased corrosion rates have been observed at very low H2S concentrations, while the corrosion rate was reduced when the H2S concentration was increased so much that an iron sulfide film was formed. The reduction was most pronounced in the temperature range 60-100 °C, the range where usually highest corrosion rate in pure CO2 environment is observed11,12. In experiments, carried out in flow loops13 and in experiments reported by Hausler14 small H2S additions were found to suppress localised corrosion. The pCO2/pH2S ratio in these experiments was between 650 and 5000. Other researchers mention ratios of 200 to 500 as the borderline between carbonate and sulfide domination15,16,17. These ratios are closer to calculated values based on thermodynamic data. The steps to be taken when moving from sweet corrosion to sour corrosion prediction are large, since the type of corrosion films, corrosion attacks and mechanisms to be predicted are very different.

5 Appendix 2

3. PREDICTION MODELS 3.1 Corrosion Prediction focusing on CO2 Corrosion A short description of the various CO2 corrosion prediction models is given below. The most referred model has been developed by Shell (de Waard et al.). The first version based on temperature and CO2 partial pressure only was published in 197518. The model has been revised several times since with correction factors for effect of pH, protective scale, oil wetting, glycol and flow velocity19, 20, 21. The 1995 version represents a best fit to a large number of flow loop data generated at IFE22. The HYDROCOR model is developed within Shell to combine corrosion and fluid flow modelling and is now Shell’s preferred tool for corrosion prediction. CO2 corrosion models are coupled to models for multiphase flow, pH calculation and iron carbonate precipitation23, 24, 25. This enables calculations of the corrosion rate over a pipeline profile. Oil wetting effects are included for crude oil systems. Protection from corrosion product films is assumed only when formation water is not present due to risk for localized attack. CASSANDRA is a model representing BP’s implementation of the de Waard models and including BP’s experience in using these models26. The spreadsheet with the model is openly available. The effect of protective corrosion films can be included or excluded by the user by choosing the scaling temperature. Oil wetting effects are not considered in Cassandra. BP has also issued detailed and practical guidelines on how the prediction model may be used for design purposes26. Important concepts are the use of corrosion inhibitor availability rather than inhibitor efficiency and the use of corrosion risk categories as a way of quantifying the corrosion risk at the design stage. The CORMED prediction tool27 developed by Elf predicts the probability of corrosion attack for wells. It is based on a detailed analysis of Elf Aquitaine’s field experience on CO2 corrosion. The model calculates in-situ pH, free acetic acid and the Ca2+/bicarbonate ratio, and then predicts the corrosivity of wells as either a low risk, medium risk or a high risk for attack. This tool identifies the amount of acetic acid as a very important parameter. The prediction tool has been developed for wells. The LIPUCOR corrosion prediction program28 calculates corrosion rates based on temperature, CO2 concentration, water chemistry, flow regime, flow velocity, characteristics of the produced fluid, and material composition. The program is developed by Total and is based on both laboratory results and a large amount of field data. The correlation with field data makes this model considerably less conservative than the worst case models. The NORSOK model29, 30 is an empirical model mainly based on laboratory data at low temperature and a combination of lab and field data at temperatures above 100 °C. The model is fitted to much of the same IFE lab data22 as the de Waard 95 model, but includes in addition more recent experiments at 100 to 150 °C3, 13. The model takes larger account for the effect of protective corrosion films and predicts therefore lower corrosion rates at high temperature and high pH than the de Waard and Hydrocor models. The model has been developed by the Norwegian oil companies Statoil, Norsk Hydro and Saga Petroleum, and has been issued as a NORSOK standard for the Norwegian oil industry. The spreadsheet with the model is openly available.

6 Appendix 2

The KSC Model is a mechanistic model for CO2 corrosion with protective corrosion films developed at IFE within the multiclient project Kjeller Sweet Corrosion V13, 31. The model is based on an electrochemical model32 by building it together with a transport model. The different electrochemical, chemical and transport processes are simulated. The properties of the protective corrosion films are correlated with a large number of loop experiments in the KSC projects. The model calculates the concentration profiles and fluxes of the different species and the resulting corrosion rate. PREDICT is a software tool developed by InterCorr (CLI International)33. The basic part of the model is based on the de Waard models, but other correction factors are used together with a so-called effective CO2 partial pressure calculated from the system pH. Predict includes strong effects of oil wetting and protective corrosion films. The corrosion model developed by OLI Systems combines a thermodynamic model for the concentration of molecular and ionic species of aqueous systems with an electrochemical corrosion model and a model for formation and dissolution of iron carbonate or sulfide scales34. The model is based on detailed thermodynamic calculation of the phase equilibria and the concentration of the different species in the system. The model includes effects of protective corrosion films, but does not take any effect of oil wetting into account. The Electronic Corrosion Engineer model developed by Intetech is based on the de Waard 95 model, but with a module for calculation of pH from the water chemistry and bicarbonate produced by corrosion, and a new oil wetting correlation, which is dependent on the oil density, the liquid flow velocity and the inclination of the flow35. SweetCor is an information system developed by Shell in the U.S. for analysis of CO2 corrosion by managing a large database of corrosion data from laboratory experiments and field data36. The approach is to group data by ranges of temperature and CO2 partial pressure or by the stable corrosion product. Statistical analysis of the grouped data is used to make correlations for predicting corrosion rates for specific conditions. The CO2 corrosion model developed at the University of Tulsa is a mechanistic single-phase flow model with detailed modelling of the kinetics of electrochemical reactions and mass transfer37. The model calculates the pH and the corrosion rate in presence iron carbonate scales. A collection of models for predicting corrosion rates in multiphase flow has been developed at Ohio University38,39. A model for oil/water flow incorporates water chemistry, mass transfer and electrochemical kinetics and is coupled with modules for calculating the solution pH, height of water and oil layers in three-phase flow and corrosion rates in slug flow. This "old Ohio model" is now being replaced with a new mechanistic based model following change of leadership in the corrosion group at Ohio University. The ULL corrosion model consists of a package of programs developed for gas condensate wells by the University of Louisiana at Lafayette (ULL), formerly the University of Southwestern Louisiana (USL)40. The model calculates temperature and pressure profiles, phase equilibria, flow rates and flow regime and then calculates the pH profile and predicts the corrosion rate profile. The model includes effects of oil wetting when hydrocarbon condensation occurs.

7 Appendix 2

Corpos is a tool developed by CorrOcean where results from multiphase flow calculations are combined with water chemistry calculation and a point corrosion model41. Corpos includes calculation of pH and a probability of water wetting and uses the Norsok CO2 corrosion model to calculate the corrosion rate in several points along the pipeline. IFE has implemented some of the CO2 corrosion models into the comprehensive three-phase fluid flow model OLGA42. The corrosion models implemented so far are the Norsok model and the de Waard 1995 models. Temperature, pressure, wall shear stress and liquid flow velocity profiles along the pipeline are calculated by the fluid flow model. This is then used to calculate CO2 partial pressure, pH and corrosion rate profiles along the pipeline. Oil wetting is assumed if the fluid flow model predicts that oil and water are dispersed with the oil as the continuous phase. 3.2 Corrosion Prediction in the Presence of H2S All the models discussed in the previous section are primarily dealing with CO2 corrosion and are not particularly suited for situations with appreciable amounts of H2S. When even small amounts of H2S are present, the corrosion products are iron sulfide rather than iron carbonate, since iron sulfide is much less soluble and precipitates much more rapidly than iron carbonate. Prediction models developed on the basis of formation of protective iron carbonate films can therefore not be used for situations where iron sulfide films are formed instead. Some of the models discussed above use the H2S content in the pH calculation without actually predicting sulfide-dominated corrosion, while some of the models give a warning that the results are not valid when the H2S content is above a certain low level, which can be as low as 1 - 10 mbar H2S. Lipucor gives a H2S corrosion rate dependent on H2S partial pressure and flow velocity in addition to the CO2 corrosion rate28. Predict gives a marked reduction in corrosion rate due to iron sulfide films when the H2S content is higher than about 1 mbar43, the OLI model indicates a reduction at even lower H2S levels44. Hydrocor uses the sweet corrosion rate multiplied with a pitting factor between 0.7 and 6 for the sulfide dominated regime defined as pCO2/pH2S < 20. This pitting factor indicates the tendency to pitting in sulfide dominated systems and increases with chloride content and presence of elemental sulfur25. The ECE model uses a sulfide scale factor, which is dependent on pH. This gives a marked reduction in corrosion rate due to iron sulfide films already around 1 mbar H2S partial pressure. The prediction of general and localized corrosion in systems with H2S in addition to CO2 is little developed and highly uncertain, and several of the models, but not all, give warnings stating either that the results are not valid above a certain H2S level or that the corrosion rates in the presence of H2S are uncertain. There is obviously a need for development of H2S corrosion models that take the different types of sulfide films into account. The steps to be taken when moving from sweet corrosion to sour corrosion prediction are large since the type of corrosion films, corrosion attacks and mechanisms to be predicted are very different. The best advice for the time being is to be very careful in using the discussed models for situations with more than trace amount of H2S. Some of the main challenges for future development of models for sour corrosion are:

8 Appendix 2

• Corrosion films: The models must be able to predict the consequences of the formation and co-existence of various types of iron sulfides in multi-layer corrosion films in order to give a correct estimate of the sour corrosion process. Potential/pH diagrams with the different iron sulfides exist45, and in order to predict film formation these data may be used to calculate gradients for electrode potentials and concentrations of reactive species across the corrosion films. Research on this field is ongoing46, but far from complete. Competitive formation between FeS and FeCO3 must be taken into account for very low H2S partial pressures, including the effect of local depletion in cases where H2S completely or partially reacts to form FeS.

• Corrosion attacks: Localised corrosion frequently occurs in sour systems. These attacks have different morphologies (pitting, crevice corrosion), but for all types the cathodic and anodic corrosion reactions are to some extent physically separated on a macroscopic scale, i.e. an anodic area can be recognised by the unaided eye. In order to predict formation and growth of anodic areas the steel surface must be mapped and the corrosion process modelled on a molecular scale. Some iron sulfide films, e.g. mackinawite, are conductive and may serve as substrates for the cathodic reactions. This phenomenon will have to be quantified and implemented in the model. In addition, quantitative knowledge of the influence of all elements in the steel microstructure on localised attack initiation is required, including grain boundaries, cementite lamellae, inclusions and contaminants. A statistical/empirical approach may be an easier way to model the distribution of anodic and cathodic reactions, but this requires a fair number of reliable data from fields and laboratory tests.

• Corrosion mechanisms: Even if localised corrosion is not expected, modelling general sour corrosion rates are not an easy task. Implementation of the different corrosion mechanisms described above, such as FeS dissolution control, mass transport control and corrosion in aqueous layer beneath films represent a major challenge.

Based on the complexity, a mechanistic model for sour corrosion seems to be beyond the level of present knowledge. It is the opinion of the authors that the most effective modelling approach for the near future seems to be a semi-empirical one, using lab and field corrosion experience as a foundation.

4. UNCERTAINTIES RELATED TO THE INPUT PARAMETERS AND THE CORROSION MECHANISMS IN CORROSION PREDICTION

4.1 Water Composition and Determination of pH One of the most crucial aspects in corrosion evaluation of oil and gas wells and pipelines is to obtain a realistic estimate of the actual pH in the water phase. For cases with only condensed water this should involve an evaluation of whether the pH of the condensed water is increased due to bicarbonate produced by corrosion. Some of the models include bicarbonate produced by corrosion in the pH calculation. When formation water is produced it is important to obtain good water analysis data, especially with respect to bicarbonate and organic acids. Lack of good quality formation water analysis represents one of the most important uncertainties in corrosion prediction for new developments. In many cases formation water samples are very scarce or not collected at all during the first well tests. Formation water samples are often contaminated with drilling or completion fluids, and this will introduce uncertainties especially for the bicarbonate content, which is essential for the pH calculations.

9 Appendix 2

Another major uncertainty is when and how much formation water will actually be produced. It is often very uncertain when formation water breakthrough will occur. The water chemistry and hence the pH and the resulting corrosion rate can change drastically if the water composition changes from condensed water only to a mixture of condensed and formation water. If only small amounts of formation water are produced the actual water chemistry can be very different from the formation water analysis, as the formation water may be strongly diluted by condensed water. After injection water breakthrough late in the field life the water chemistry can change considerably again. The reported pH in a water analysis is often useless for a corrosion prediction, as it is usually measured at atmospheric conditions after depressurization. This gives no information about the actual pH in the pipeline, which must be calculated from the CO2 partial pressure, temperature, bicarbonate content in the water and ionic strength. Several of the models have pH modules that perform such calculations. On the other hand, some of the models show little dependence of pH on the corrosion rate, and therefore uncertainties in the pH calculations will not have a large effect on the predicted corrosion rate. In addition to the CO2/bicarbonate buffer system also the H2S/sulfide and the acetic acid/acetate buffering systems can be important for determining the actual pH value. The presence of acetic acid and other organic acids can have a very strong impact on corrosion rates, especially at low CO2 partial pressures and for cases when top-of-line corrosion is a concern. Most of the models do not take this into account. Presence of organic acids can give too high values for bicarbonate and hence too high calculated pH values if organic acids are not measured in the water analysis47. For old water samples it is very rare to find measurements of organic acids. In some cases the specified water chemistry from a water analysis can indicate supersaturation of calcium carbonate. It may be advisable to check the water analysis for supersaturation of calcium carbonate at reservoir conditions, which may indicate an erroneous bicarbonate analysis since supersaturation of calcium carbonate is not possible in the reservoir. In this case the bicarbonate value may be adjusted to the bicarbonate solubility at reservoir conditions with the actual calcium content. 4.2 Effect of Protective Corrosion Films in CO2 Environment One of the difficulties with prediction of CO2 corrosion of carbon steel is the very important effect of protective iron carbonate films especially at high temperature or high pH. At low temperature the iron carbonate solubility is high and the precipitation rate slow, and protective films will therefore not form unless the pH is artificially increased. At high temperature the iron carbonate solubility is lower and the precipitation rate much faster, and very dense and protective iron carbonate films can form. This can lower the corrosion rate from several mm/y for carbon steel without any corrosion films to less than 0.1 mm/y when protective films are present. The effect of protective corrosion films can almost be considered as an on/off switch, and the success of a prediction model depends to a large degree on whether it is able to predict the presence or absence of protective films or localized attack reliably, rather than the ability to predict the general corrosion rate with a certain accuracy. The picture is further complicated by the susceptibility to form localized corrosion attacks in the form of pits or mesa attack on steel surfaces with a partially protective film. As the formation of stable films depends on the FeCO3 supersaturation, the prediction becomes very dependent on accurate information about the water composition and the amount of corrosive gases. The accuracy in the prediction will therefore usually be limited by the quality of the input data.

10 Appendix 2

The effect of protective corrosion films on the predicted corrosion rate varies considerably between the models. Some of the models include very strong effects of protective films, while others include only moderate effects of protective films or do not take into account protective corrosion films for formation water cases due to high risk for localized attack for such cases. Generally the models with the strongest effects of protective films rely on easy formation of corrosion films with good protective properties and absence of localized attack, while the models with weak effects of protective corrosion films assume that the films have only limited protectiveness or that there is a high risk for localized attack. The de Waard, Cassandra, Hydrocor, Sweetcor and ECE models have only weak effects of protective corrosion films. The Norsok, Corpos, Lipucor and KSC models have moderate effects of corrosion films, while the Tulsa and Predict models have very strong effects of protective corrosion films. The worst case CO2 corrosion rate occurs when protective iron carbonate films are not formed. Typical worst case conditions are 1 bar CO2 and pure water at ambient temperature, which give a pH value around 4 and no protective films. This is relatively easy to predict and the models perform similar. The differences between the models become evident at higher temperatures or higher pH, when protective iron carbonate films start to form. This can be illustrated by two field data examples from the joint industry project at Institute for Energy Technology where the different models were evaluated. The first example is a short onshore gas line with 1.2 bar CO2 partial pressure and only condensed water at a temperature of about 50 °C. Under these conditions protective films are not expected to form, and a corrosion rate of 4 mm/y was measured by ultrasonic thickness measurements. Most of the models calculated a pH around 4 for this case, and the predicted corrosion rates varied from 4 to 13 mm/y. It is not important that some of the models predicted higher corrosion rates than actually observed, as long as none of the models failed to predict that these conditions would give unacceptably high corrosion rates. The second example is an oil well with 1.6 bar CO2 partial pressure and temperature from 110 °C downhole to 70 °C at the wellhead. Corrosion model calculations gave pH values between 5 and 5.5 for the formation water chemistry in this well. This well suffered penetrating mesa attack corresponding to 5 mm/y corrosion rate close to the top of the well, while calliper readings indicated around 0.5 mm/y corrosion rate further down in the well. Figure 5 shows the measured corrosion rates and the corrosion rates predicted by some of the models in Table 1. Model A give much credit for effects of oil wetting and protective corrosion films. This model predicts the low corrosion rates downhole quite well, but it is unable to predict the failure close to the wellhead. Model B gives little credit for oil wetting and protective corrosion films. This model is able to predict the high corrosion rate leading to the failure, but it does not predict the low corrosion rates further down in the well. Model C gives more credit for protective corrosion films at high temperature, and predicts highest corrosion rate close to the wellhead where the failure occurred. The models behaved differently for other field cases, and none of them are able to give reliable predictions for all cases.

11 Appendix 2

4.3 Effect of Oil Wetting It is important to know whether water or oil wets the steel surface since corrosion takes place only when water is present at the surface. If the water is transported as a water-in-oil emulsion or dispersion corrosion can be substantially reduced. The degree of oil wetting depends heavily on flow conditions, water cut and the properties of the actual hydrocarbon, and this is difficult to assess without performing laboratory or field tests with the actual oil-water mixture. Some of the models have a very strong effect of oil wetting for some flow conditions, while other models do not consider oil wetting effects at all, either because oil wetting effects are not modeled or because it is believed that water will wet the steel surface and cause corrosion somewhere in the pipeline anyway. The Norsok, Cassandra, Cormed, Sweetcor, Tulsa and KSC models have no effects of oil wetting. In the simplest form oil wetting is treated as an on/off switch in the models: either the surface is water wet with full corrosion or oil wet with zero corrosion. In the de Waard model oil wetting and zero corrosion is assumed if the liquid is larger than 1 m/s and the water cut is less than 30 %. This effect is used for crude oil only and not for condensate, as water is considered to separate out much more easily in condensate systems19, 20. In the more recent Hydrocor model these limits are changed to 40 % and 1.5 m/s, and partial protection is assumed when the water cut and flow velocity are above these limits. Also in the Hydrocor model no protection is assumed for condensate systems25. Another recently developed model is the ECE model, where the pipeline corrosion is related to different modes of water entrainment35. Depending on conditions of water cut, deviation angle and velocity the fluid can consist of either a water in oil emulsion (mode I) or a water phase separated from the oil phase (mode II and III). A water in oil emulsion will separate into two phases at a critical water cut. The resulting water phase can remain stationary at certain locations (mode II), or it may move with the fluid wetting the steel intermittently (mode II). Furthermore a link between API gravity, emulsion stability and water wetting is provided by considering changes in interfacial tensions in oil-water systems. An empirical formula gives a satisfactory description for two different oil fields of the influence of the API gravity of the oil and its water cut on the corrosion. In the model, oil wetting and low corrosion is assumed for water cut lower than the emulsion breakpoint, which is calculated from the oil density and can be up to 50 % for heavy crude oils and close to zero for light condensate. The critical flow velocity for water dropout is taken as 1 m/s for horizontal flow and lower for inclined flow. The Lipucor model includes strong effects of oil wetting. The critical velocity for water entrainment in the oil is often calculated to around 0.5 m/s, and for liquid velocities above this low corrosion rates are often predicted. The Predict model also includes strong effects of oil wetting33. The ULL model for gas wells has a strong effect of oil wetting when hydrocarbon condensation occurs. It typically predicts oil wetting and no corrosion for the part of the well where hydrocarbon condensation occurs, and corrosion when only water condenses. In the corrosion module in the OLGA fluid flow model the transition between separated and dispersed water/oil in the fluid flow model is used to determine the degree of water wetting42. Oil wetting is assumed if oil and water are dispersed and the water cut is less than 30 %, giving a dispersion with the oil as the continuous phase. Many of the corrosion models do not take into account the effect of oil and they are therefore more conservative and predict corrosion rates that often prohibit the use of carbon steel under conditions where it is likely that carbon steel could be applied. Many of the companies however, are reluctant to give any credit for entrainment of the oil as experiences from some fields have shown that corrosion can take place at water cuts as low as 1 %. Better understanding of the parameters controlling the

12 Appendix 2

entrainment of oil and the oil wetting mechanism has a potential to improve the predictions very much and increase the confidence in the models. 4.4 Top-of-Line Corrosion The discussion above concerns primarily corrosion in the bulk water phase. The situation is quite different for top-of-line corrosion when water condenses out in the upper part of a pipeline. The condensing water is unbuffered with low pH, but can become rapidly saturated or supersaturated with corrosion products, giving rise to increased pH and possibility for iron carbonate film formation. The top-of-line corrosion rate then becomes dependent on the water condensation rate and the amount of iron which can be dissolved in the condensing water48. The de Waard model includes a very simplified model for top-of-line corrosion19, but little attention was paid to top-of-line corrosion until recently, when top-of-line corrosion problems were encountered in a few wet gas pipelines. Top-of-line corrosion has been now included in the Lipucor and Hydrocor models49, 50 and in the OLGA corrosion module. Top-of-line corrosion is primarily a concern in the first few kilometres of wet gas pipelines with relatively high inlet temperatures, as the water condensation rate is rapidly reduced when the temperature decreases. The presence of acetic acid in the gas may increase the top-of-line corrosion rate considerably, as it increases the amount of iron which can be dissolved in the condensing water before protective corrosion films are formed. The few cases where top-of-line corrosion problems have been reported have been associated either with unusually high pipeline cooling rates as in river crossings or with presence of acetic acid in the gas.

5. PARAMETERS USUALLY NOT INCLUDED IN CORROSION PREDICTION MODELS

5.1 History and Operational Parameters

Protective film formation is accelerated by all measures restricting the transport of reaction products from the surface. Freshly ground and continuously exposed steel is therefore much more susceptible to high corrosion rates than real pipelines, where the surface is covered by oxide layers grown during rolling and storage and carbonate layers formed during stagnant periods and periods with low replenish rate of water. These effects have been studied extensively3 in CO2 environment and examples of results obtained at 20°C are shown in Figure 6. It is seen that the "dry"∗ and stagnant periods trigged the formation of protective films and that the corrosion rates were reduced to less than 0.1 mm/year shortly after the incidents. Corrosion rates on the pre-oxidised specimens remained low all the time. It is seen in many corrosion experiments that protective films do not form at all before the system has gone through a shut down giving stagnant conditions or reduced water wetting of the surface. Without taking the operational incidents and the history of the steel pipe into consideration, a mechanistic corrosion prediction model will over predict the actual corrosion rate in the field. It is difficult to model the described effect. The likelihood for shutdowns, stagnant periods, etc. is high however and the information can be used qualitatively as an on/off protection effect. This approach has

* Period where the steel surface is water wet, but the bulk water is removed such that only a thin film covers the steel surface. The result is no replenishment of the water and that gives accumulation of corrosion product in the film

13 Appendix 2

been used for the qualification of pH-stabilised systems. When corrosion models are calibrated against field data the effect might be picked up and be partly included in the models. 5.2 Effect of Microstructure and Trace Elements on Protective Corrosion Product Films The microstructure and the concentration of the trace elements in the steel and the weld respectively will be different and the differences can also be large between different batches and types of carbon steel, even though they have the same designation. It has been seen that these differences can have a large effect on the corrosion rate. This is attributed to changes in the growth of corrosion product on the surface and to interaction between the corrosion film and natural and added inhibitors. Iron carbonate layers can form on steel surfaces at all temperatures if the supersaturation of iron carbonate in the bulk is sufficiently high. High supersaturation, however, is not the only requirement for successful protection of the steel. In order to obtain satisfactory protection, the film must be adherent and cover the whole surface as mesa attack will develop if parts of the film break down and do not reform3. Both lab experiments and field experience have shown that the protective properties and the adherence of the corrosion product film can vary much for carbon steels with apparently the same composition and microstructure. There is however, no agreement about the mechanism and how composition and microstructure affect the growth and stability of the iron carbonate film. It has been suggested that normalized plain steel with a pearlitic microstructure is superior to the hardened and tempered alloy steel with a martensitic microstructure under the same conditions, but the opposite has also been reported51-55. Figure 7 shows the corrosion rate and the area of localized attack on a low alloyed carbon steel that has been given various heat treatments56. The heat treatment included hardening, tempering and spheroidizing of the carbide phase. The steel samples were tested in a corrosion flow loop under conditions where protective film can form. The corrosion rate in the deepest attack and the adherence of protective corrosion product films decreased with increasing tempering temperature. No protective film formed on the ferritic pearlitic (as received steel) and the spheroidized steel coupons with the largest particles of carbide. Only one test sample obtained full protection and a low corrosion rate. It had been austenitized, quenched and tempered for 60 min. at 300°C and was exposed at 4.1 m/s. Two other steels with 0.53 and 0.87 % Cr respectively were also heat treated and exposed in the same test. Corrosion films with protective properties formed on all the heat-treated chromium containing steel coupons and no mesa attack was observed. The results clearly show that a range of quite different corrosion rates and types of attack can be obtained in the same environment when the microstructure or the amounts of trace alloying elements vary. When these effects are not included in the models, it must obviously affect the accuracy of the predicted corrosion rate and the type of attack. 5.3 Effect of Microstructure and Trace Elements on Inhibition Inhibitors do not always perform as intended in the field57-60. The reasons for such problems are complex and not yet well understood. Laboratory experience shows that inhibitors perform differently on corroded surfaces and freshly ground surfaces. In the field, inhibitors encounter steel surfaces that are covered with different kinds of corrosion products, such as mill scale from pipe production and rust from storage and pipeline testing. Furthermore, a pipeline may have been in operation for several years

14 Appendix 2

before increasing water cut necessitates inhibition. In this case the metal surface might be covered with iron carbonate precipitates and sulfides, uncorroded iron carbide, and different types of scale. These products may significantly affect the performance of the inhibitor. Inhibitor tests have shown that long term precorrosion strongly affected the performance of the tested inhibitors61. In general, the inhibitor efficiency decreased with increasing precorrosion time and large differences in performance were found between different steels and different inhibitor products as shown in Figure 862. The effect of steel structure and composition was tested on 16 different steels using various commercial inhibitor products. The results show that the inhibitor efficiency varied from 0 % to almost 100 %. The ranking between the steels was almost the same when they were tested with three different inhibitors, and the inhibition problem was mainly connected to the cathodic inhibition. The large differences observed between different steels emphasize that inhibitor testing and corrosion monitoring should be carried out with steel specimens that closely resemble the actual pipeline steel. It also emphasizes the importance to include the steel properties when the corrosion risk is evaluated. It can be concluded that predicting corrosion rates for inhibited systems where only the properties of the environment is taken into consideration can give serious under-prediction.

5.4 Effect of Crude Oils containing Natural Inhibitors

As discussed previously, the entrainment of water in the crude oil can contribute significantly to the mitigation of internal corrosion. It has been much less focus on how small quantities of organic compounds containing sulphur, oxygen, nitrogen and even trace amounts of metals in the crude can affect the wettability. Field experience has shown that the actual corrosion rate can be low despite prediction of a free water phase and a high corrosivity. There are reports claiming that even 1 to 5% crude oil give moderate corrosion protection for CO2 corrosion63-66. The cause for this behaviour is most probably that certain components in the crude oil adsorb to the steel surface in a way similar to how commercial corrosion inhibitors do. These "natural corrosion inhibitors" may affect the electrochemical, chemical or mass transfer reactions taking place in the corrosion process or they may act as surfactants changing the wettability of the surface.

On the other hand, crude oil may also contain species that increase the corrosivity like acetic acid (or other related weak acids)47, 67. Literature data show that the acetic acid concentration in the oil is 1-2 % of the acetic acid concentration in the water68. This means that the oil may act as a reservoir for acetic acid in systems with low watercut. Little systematic work has been done to identify the critical factors for the protective effect related to surface-active components in the oil (natural inhibition) and to preferential oil wetting of the pipe wall at high water cut. Better understanding of these effects may give a large improvement in prediction, design and operation of pipelines

15 Appendix 2

6. CONCLUSIONS

6.1 CO2 Models

Several prediction models for CO2 corrosion of oil and gas pipelines are available. The models have very different approaches in accounting for oil wetting and the effect of protective corrosion films, and this accounts for much of the differences in behaviour between the models. Some of the models have a very strong effect of oil wetting for some flow conditions, while other models do not consider oil wetting effects at all. Some models include strong effects of protective iron carbonate films especially at high pH or high temperature. These models rely on easy formation of protective corrosion films and absence of localized attack, while the models with weak effects of protective corrosion films assume that the films have only limited protectiveness or that there is a high risk for localized attack. For field case failures encountered at low pH and moderate temperature all the models are capable of predicting a high corrosion. The variation in predicted corrosion rates and the deviation from the corrosion rates observed in the field becomes much larger for systems at high temperature and high pH, where protective corrosion films may form, and for systems where some of the models assume oil wetting. In general, no models have been able to predict the corrosion attack with a high accuracy for the majority of the field cases. Models that generally predict high corrosion rates have of course a better fit when the failure cases are evaluated, but these models also predict high corrosion rates for system that perform well. If these models had been used for design they would most probably not support the use of carbon steel in many of these cases. 6.2 H2S Models The prediction of general and localized corrosion in systems with H2S in addition to CO2 is little developed and highly uncertain, and several of the CO2 corrosion models, but not all, give warnings stating either that the results are not valid above a certain H2S level or that the corrosion rates in the presence of H2S are uncertain. There is obviously a need for development of H2S corrosion models that take the different types of sulfide films into account. Based on the complexity, a mechanistic model for sour corrosion seems to be beyond the level of present knowledge. It is the opinion of the authors that the most effective modelling approach for the near future seems to be a semi-empirical one, using lab and field corrosion experience as a foundation. The steps to be taken when moving from sweet corrosion to sour corrosion prediction are large since the type of corrosion films, corrosion attacks and mechanisms to be predicted are very different. The best advice for the time being is to be very careful in using the discussed models for situations with more than a few mbar H2S. 6.3 Limitation of Input Parameters The accuracy of the corrosion rate prediction is not only limited by the quality of the prediction models. The bottleneck in the prediction can rather be the quality of the input data and the understanding of the corrosion controlling mechanisms.

16 Appendix 2

Input data that are recognized to have large uncertainties, particularly in the design phase includes:

• the concentration of organic acids. It is difficult to get good samples and organic acids are often not measured at all.

• the species in the water that control pH and precipitation. These species include salt content, salt composition, in particular bicarbonate content and amount of dissolved CO2, H2S and acetic acid. All these data depend very much on the quality of the water samples collected under test drilling.

• flow related parameters that control water entrainment. Parameters like water cut, flow velocity, flow regime and GOR will vary during the field life and is difficult to forecast with high accuracy.

• components that affect wettability. They can be natural occurring surface active components and added chemicals.

• interaction of the corrosion inhibitors with corrosion product on the steel surface, fines in the liquid and the surface of emulsions.

6.4 Challenges for Future Model Work

Old models are continuously improved and new models with increased complexity are coming up and being offered to the operators. It is, however, still a problem that the models apparently do not predict the actual field corrosion rate with high confidence. This can partly be attributed to the uncertainties in the input parameters, but as important is the lack of understanding and the problem that the models do not take all the corrosion controlling parameters/mechanisms into proper account. Important parameters that are not usually treated to any extent in the models that have been analysed are:

• Effect of history and operational parameters - Freshly ground and continuously exposed steel is much more susceptible to high corrosion rates than real pipelines, where the surface is covered by oxide layers grown during rolling and storage and carbonate layers formed during stagnant periods and periods with low replenish rate of water. In some cases protective films do not form at all before the system has gone through a shut down giving stagnant conditions or reduced water wetting of the surface. Without taking the operational incidents and the history of the steel pipe into consideration, corrosion prediction model will usually over predict the actual corrosion rate in the field.

• Effect of crude oils containing natural inhibitors - Crude oil is a complex mixture of liquid and

solid hydrocarbons with small quantities of organic compounds containing sulfur, oxygen, nitrogen and even trace amounts of metals. Despite prediction of a free water phase, these components can adsorb to the steel surface in a way similar to how commercial corrosion inhibitors do and thus reduce corrosion by changing the wettability of the surface.

• Effect of microstructure and concentration of the trace elements - Significant variation can be

seen in the mother material and the welds and between different batches of mother materials. The consequences are large differences in the corrosion rates and type of attacks. This is

17 Appendix 2

attributed to changes in the growth of corrosion product on the surface and to interaction between the corrosion film and natural and added inhibitors.

REFERENCES

1. Rolf Nyborg, "Overview of CO2 Corrosion Models for Wells and Pipelines" CORROSION/2002, Paper No. 2233. NACE International, Houston, 2002.

2. A. Dugstad, "The importance of FeCO3 supersaturation on the CO2 corrosion of carbon steels", CORROSION/92, Paper No. 14. NACE, 1992.

3. R. Nyborg, A. Dugstad " Understanding and Prediction of Mesa Corrosion Attack", CORROSION/2003, Paper No. 3642, NACE International, Houston, 2003

4. K. Videm, A. Dugstad, "Effect of Flow Rate, pH, Fe2+ Concentration and Steel Quality on the CO2 Corrosion of Carbon Steels", CORROSION/87, Paper No. 42. NACE, 1987.

5. K. D. T. Rickard, " The Chemistry of Iron Sulfide Formation at Low Temperatures ", p. 67-95, Stockholm Contr. to Geology Vol. 20, 1969.

6. D. W. Shoesmith, P. Taylor, M. G. Bailey, D. G. Owen, " The Formation of Ferrous Monosulfide Polymorphs during the Corrosion of Iron by Aqueous Hydrogen Sulfide at 21 °C", Journal of Electrochemichal Society, Vol. 127, No. 5, p. 1007-1015, May 1980.

7. P.H. Tewari, A.B. Campbell, "Dissolution of Iron During the Initial Corrosion of Carbon Steel in Aqueous H2S Solutions", Can. J. Chem. Vol. 57, p. 188-196, 1979.

8. J. Kvarekvål, "The influence of small amounts of H2S on CO2 corrosion of iron and carbon steel", Eurocorr 97, European Federation of Corrosion, Event No. 208, Trondheim, Norway, 22-25 September 1997.

9. P.R. Rhodes, "Corrosion mechanism of carbon steel ", Extended abstracts, Electrochemical Society. No. 2, Vol. 76, 1969.

10. N.N. Bich, K. Goerz, "Caroline Pipeline Failure: Findings on Corrosion Mechanisms in Wet Sour Gas Systems Containing Significant CO2", CORROSION/96, Paper No. 26, NACE International, Houston, 1996.

11. A. Ikeda, M. Ueda, S. Mukai, "Influence of Environmental Factors on Corrosion in CO2 Source Well", CORROSION/85, Paper No. 29, NACE, Houston, 1985.

12. M. Ueda, A. Ikeda, "Effect of Microstructure and Cr Content in Steel on CO2 Corrosion", CORROSION/96, Paper No. 13, NACE International, Houston, 1996.

13. R. Nyborg, "Kjeller Sweet Corrosion V, Final report", Report no. IFE/KR/F-98/075, Institute for Energy Technology, 1998.

14. R.H. Hausler, D.W. Stegmann, C.I. Cruz, D. Tjandroso, "Laboratory Studies on Flow Induced Localized Corrosion in CO2/H2S Environments; II. Parametric Study on the Effects of H2S, Condensate, Metallurgy, and Flowrate.", CORROSION/90, Paper No. 6, NACE, Houston, 1990.

15. J.A. Dougherty, "Factors Effecting H2S and H2S/CO2 Attack on Carbon Steels under Deep Hot Well Conditions", CORROSION/88, Paper No. 190, NACE, Houston, 1988.

16. A.K. Dunlop, H.L. Hassell, P.R. Rhodes, "Fundamental Considerations in Sweet Gas Well Corrosion", CORROSION/83, Paper No. 46, NACE, Houston, 1983.

18 Appendix 2

17. M.J.J. Simon-Thomas, J.C. Loyless, "CO2 Corrosion in Gas Lifted Oil Production - Correlation of Predictions and Field Experience", CORROSION/93, Paper No. 79, NACE International, Houston, 1993.

18. C. de Waard, D.E. Milliams, "Prediction of Carbonic Acid Corrosion in Natural Gas Pipelines", 1st International Conference on the Internal and External Protection of Pipes, Paper F1, Univ. of Durham, UK, 1975.

19 . C. de Waard, U. Lotz, D.E. Milliams, "Predictive Model for CO2 Corrosion Engineering in Wet Natural Gas Pipelines", Corrosion, Vol. 47, No. 12, p. 976, 1991.

20. C. de Waard, U. Lotz, "Prediction of CO2 Corrosion of Carbon Steel", CORROSION/93, Paper No. 69. NACE, 1993.

21. C. de Waard, U. Lotz, A. Dugstad, "Influence of Liquid Flow Velocity on CO2 Corrosion: A Semi-Empirical Model", CORROSION/95, Paper No. 128. NACE, 1995.

22. A. Dugstad, L. Lunde, K. Videm, "Parametric Study of CO2 Corrosion of Carbon Steel", CORROSION/94, Paper No. 14. NACE International, 1994.

23. B.F.M. Pots, "Mechanistic Models for the Prediction of CO2 Corrosion Rates under Multi-Phase Flow Conditions", CORROSION/95, Paper No. 137. NACE International, 1995.

24. E.W.J. van Hunnik, B.F.M. Pots, E.L.J.A. Hendriksen, "The formation of Protective FeCO3 Corrosion Product Layers in CO2 Corrosion", CORROSION/96, Paper No. 6. NACE International, 1996.

25. B.F.M. Pots, R.C. John, I.J. Rippon, M.J.J.S. Thomas, S.D. Kapusta, M.M. Girgis, T. Whitham, "Improvements on de Waard - Milliams Corrosion Prediction and Applications to Corrosion Management", CORROSION/2002, Paper No. 02235, NACE International, Houston, 2002.

26. A.J. McMahon, D.M.E. Paisley, "Corrosion Prediction modelling - A Guide to the Use of Corrosion Prediction Models for Risk Assessment in Oil and Gas Production and Transportation Facilities", Report No. ESR.96.ER.066, BP International, Sunbury, 1997.

27. J.L. Crolet, M.R. Bonis, "Prediction of the Risks of CO2 Corrosion in Oil and Gas Well", SPE Production Engineering, Vol. 6, No. 4, p. 449, 1991.

28. Y.M. Gunaltun, "Combining research and field data for corrosion rate prediction", CORROSION/96, Paper No. 27. NACE International, 1996.

29. A.M.K. Halvorsen, T. Søntvedt, "CO2 Corrosion Model for Carbon Steel Including a Wall Shear Stress Model for Multiphase Flow and Limits for Production Rate to Avoid Mesa Attack", CORROSION/99, Paper No. 42. NACE International, 1999.

30. "CO2 Corrosion Rate Calculation Model", NORSOK standard No. M-506, Norwegian Technology Standards Institution, http://www.nts.no/norsok, June 1998.

31. S. Nesic, M. Nordsveen, R. Nyborg, A. Stangeland, "A Mechanistic Model for CO2 Corrosion with Protective Iron Carbonate Films", CORROSION/2001, Paper No. 40. NACE International, 2001.

32. S. Nesic, J. Postlethwaite, S. Olsen, "An Electrochemical Model for Prediction of Corrosion of Mild Steel in Aqueous Carbon Dioxide Solutions", Corrosion, Vol. 52, No. 4, p. 280, 1996.

33. S. Srinivasan, R.D. Kane, "Prediction of Corrosivity of CO2 / H2S Production Environments", CORROSION/96, Paper No. 11, NACE International, Houston, 1996.

34. A. Anderko, R.D. Young, "Simulation of CO2/H2S Corrosion Using Thermodynamic and Electrochemical Models", CORROSION/99, Paper No. 31, NACE International, 1999.

19 Appendix 2

35. C. de Waard, L. Smith, B.D. Craig, "The Influence of Crude Oil on Well Tubing Corrosion Rates", CORROSION/2003, Paper no. 3629. NACE International, 2003.

36. R.C. John et al, "SweetCor: An Information System for the Analysis of Corrosion of Steels by Water and Carbon Dioxide", CORROSION/98, Paper No. 20. NACE International, 1998.

37. E. Dayalan et al, "CO2 Corrosion Prediction in Pipe Flow under FeCO3 Scale-Forming Conditions", CORROSION/98, Paper No. 51. NACE International, 1998.

38. W.P. Jepson et al, "Model for Sweet Corrosion in Horizontal Multiphase Slug Flow", CORROSION/97, Paper No. 11. NACE International, 1997.

39. R. Zhang, M. Gopal, W.P. Jepson, "Development of a Mechanistic Model for Predicting Corrosion Rate in Multiphase Oil/Water/Gas Flows", CORROSION/97, Paper No. 601. NACE International, 1997.

40. Cedric D. Adams, James D. Garber, Rakesh K. Singh, "Computer modelling to Predict Corrosion Rates in Gas Condensate Wells Containing CO2", CORROSION/96, Paper No. 31. NACE International, 1996.

41. P.O. Gartland, J.E. Salomonsen, "A Pipeline Integrity Management Strategy Based on Multiphase Fluid Flow and Corrosion Modelling", CORROSION/99, Paper No. 622. NACE International, 1999.

42. R. Nyborg, P. Andersson, M. Nordsveen, "Implementation of CO2 Corrosion Models in a Three-Phase Fluid Flow Model", CORROSION/2000, Paper No. 48. NACE International, 2000.

43. S. Srinivasan, S. Tebbal, "Critical Factors in Predicting CO2/H2S Corrosion in Multiphase Systems", CORROSION/98, Paper No. 38, NACE International, 1998.

44. A. Anderko, "Simulation of FeCO3/FeS Scale Formation Using Thermodynamic and Electrochemical Models", CORROSION/2000, Paper No. 102, NACE International, 2000.

45. A. Pourbaix, M. Amalhay, A.K. Singh, "Contribution to the Mechanisms of Localised Corrosion of C-steel and 13% Cr Steel in H2S Environment", EUROCORR ’97, Trondheim, Norway, 1997.

46. J. Kvarekvål, R. Nyborg, H. Choi, "Formation of Multilayer Iron Sulfide Films During High Temperature CO2/H2S Corrosion of Carbon Steel", Paper No. 03339, CORROSION/2003, NACE International, 2003.

47. B. Hedges, L. McVeigh, "The Role of Acetate in CO2 Corrosion: the Double Whammy", CORROSION/99, Paper No. 21, NACE International, 1999.

48. S. Olsen, A. Dugstad, "Corrosion under Dewing Conditions", CORROSION/91, Paper No. 472, NACE, 1991.

49. Y. M. Gunaltun, D. Larrey, "Correlation of Cases of Top of Line Corrosion with Calculated Water Condensation Rates", CORROSION/2000, Paper No. 71, NACE International, 2000.

50. B. F. M. Pots, E. L. J. A. Hendriksen, "CO2 Corrosion under Scaling Conditions - The Special Case of Top-of-Line Corrosion in Wet Gas Pipelines", CORROSION/2000, Paper No. 31, NACE International, 2000.

51. R.W. Manuel, "Effects of Carbide Structure on the Corrosion Resistance of Steel," Corrosion J. 3, 9 (1947): p 197.

52. C.A. Palacios, J.R. Shadley, "Characteristics of Corrosion Scales on Steel in a CO2-Saturated NaCl Brine," Corrosion J. 47, 2 (1991): p 122.

20 Appendix 2

53. B. Mishra, S. Al-Hassan, D.L. Olson, M.M. Salama, "Prediction of Microstructural Effects on Corrosion of Linepipe Steels in CO2 - Brine Solution". CORROSION/93, paper No. 90, NACE, 1993.

54. G. B. Chitwood, W. R. Coyle, R. L. Hilts, "A case-history analysis of using plain carbon & alloy steel for completion equipment in CO2 service," CORROSION/94, paper No. 20, NACE, 1994.

55. D.W. Stegman et al., "Laboratory Studies on Flow Induced Localized Corrosion in CO2/H2S Environments, I. Development of Test Methodology," CORROSION/90, paper No. 5, NACE, 1990.

56. A. Dugstad, H. Hemmer, M. Seiersten, "Effect of Steel Microstructure on Corrosion Rate and Protective Iron Carbonate Film Formation", CORROSION Vol. 57, No 4.

57. D.F.Ho-Chung-Qui, A.I.Williamson, "Corrosion Experience and Inhibition Practices in Wet Sour Gas Gathering Systems", CORROSION 87, paper no. 46, NACE International, 1987.

58. M.R.Bonis, J.-L.Crolet, "An Inhibition Policy Based on Laboratory and Field Experience", CORROSION 98, paper no. 103, NACE International, 1998.

59. D.Blumer, N.Grahmann, "Continuous Corrosion Inhibitor for High Velocity Multiphase Pipelines", Proc. 8th Int. Oil Field Chemical Symposium, Paper 14, Norwegian Society of Chartered Engineers, Oslo, 1997

60. A.S. Green, B.V. Johnson, H. Choi, "Flow-Related Corrosion in Large-Diameter Multiphase Flowlines", Proc. 65th Annual Technical Conference of the Society of Petroleum Engineers, Paper SPE 20685, SPE, 1990

61. E. Gulbrandsen, S. Nesic, A. Stangeland, T. Burchardt, B. Sundfær, S.M. Hesjevik, S. Skjerve, "Effect of Precorrosion on the Performance of Inhibitors for CO2 Corrosion of Carbon Steel", CORROSION 98, paper no. 13, NACE International, 1998.

62. E.Gulbrandsen, R. Nyborg, "Effect of Steel Microstructure and Composition on Inhibition of CO2 Corrosion", CORROSION 2000, paper no. 23, NACE International, 2000.

63. M. Castillo, H. Rincon, S. Duplat, J. Vera, E. Baron: "Protective Properties of Crude Oils in CO2 and H2S Corrosion", CORROSION/2000, Paper no. 5, NACE International, 2000.

64. C. Méndez, S. Duplat, S. Hernández, J.Vera: "On the Mechanism of Corrosion Inhibition by Crude Oils", CORROSION/2001, Paper no. 30, NACE International, 2001.

65. S.E. Hernandez, S. Duplat, J.R. Vera, E. Baron: "A Statistical Approach for analyzing the Inhibiting Effect of Different Types of Crude Oil in CO2 Corrosion of Carbon Steel", CORROSION/2002, Paper no. 2293, NACE International, 2002.

66. S.E. Hernandez, J. Bruzual, F. Lopez-Linares, J.G. Luzon: "Isolation of Potential Corrosion Inhibiting Compounds in Crude Oils", CORROSION/2003, Paper no. 3330, NACE International, 2003.

67. M.R. Bonis, J.L. Crolet: "Basics of the Risk of CO2 Corrosion in Oil and Gas Wells", CORROSION/1989, Paper no. 466, NACE, 1989.

68. M.A. Reinsel, J.J. Borkowski, J.T. Sears: "Partitioning Coefficients for Acetic, Propionic, and Butyric Acids in a Crude Oil/Water System", J. Chem, Eng. Data, 39, 513, 1994.

21 Appendix 2

Table 1 - Application Limits for Different Corrosion Models

T, °C P, bar pCO2, bar pH Min Max Max Min Max Min Max de Waard 1 0 140 10 HYDROCOR 0 150 200 20 Cassandra 98 2 140 200 10 NORSOK 3 20 150 1000 10 3.5 6.5 CORMED 4 120 LIPUCOR 20 150 250 50 KSC Model 5 5 150 200 0.1 20 3.5 7 Tulsa model 6 38 116 17 PREDICT 7 20 200 100 2.5 7 Ohio model 8 10 110 20 SweetCor 5 121 0.2 170 ECE model 9 0 140 200 20 OLI model 10 120 1500 20

1 None of the de Waard papers give application limits. Max. values in nomogram shown. 2 Accepts input outside these values but displays a warning. 3 Wall shear stress between 1 and 150 Pa. Will be extended down to 5 °C. 4 Cormed accepts higher temperatures and ionic strengths but displays a warning,

as the pH calculation becomes uncertain. The corrosion risk prediction is still valid. 5 Flow velocity between 0.2 and 30 m/s. 6 Recommends these limits, but accepts input outside these values. 7 Predict does not give any limits, neither in the software nor in the manual. 8 Min. 10 % water cut. Can be used at higher pressures with fugacity coefficient input. 9 Max. CO2 and H2S content 10 % of gas composition. 10 The OLI model does not give any temperature or CO2 limits for the corrosion model.

The values are taken from the data used for the correlations in the corrosion model.

22 Appendix 2

Figure 1 - Cross Section of Carbon Steel exposed at 2 bar CO2 Partial Pressure a) Porous Fe3C film formed at 20 °C . b) Porous film filled with protective iron carbonate

formed at 80 °C. c) Steel surface partly covered with protective film

formed at 80 °C. Mesa attack occurred.

23 Appendix 2

Figure 2 - Schematic Illustration of the Effect of Various Parameters on CO2 Corrosion of Carbon Steel

24 Appendix 2

FIGURE 3b - Proposed mechanism for

growth of mesa attack. Figure 3a - Development of mesa attack as observed by in-situ video recording. The area shown is 10 x 14 mm. 3

25 Appendix 2

Pyrite Marcasite FeS2 FeS2

Cubic Orthorhombic

Figure 4 - Summary of the major iron sulfide inter-relationships in aqueous solutions and

their biogenic counterparts, modified from Rickard5.

0

1

2

3

4

5

050010001500

Depth / m

Cor

rosi

on ra

te /

(mm

/y)

Model A

Model B

Model C

Measured corrosion

Figure 5 - Example of model runs for an oil well with penetrating mesa attack close to the

wellhead.

Smythite Fe3S4

Iron Mackinawite Greigite Fe2+ FeS1-x Fe3S4

Aqueous Tetragonal Cubic

Siderite Pyrrhotite FeCO3 Fe1-xS

Ni As type

26 Appendix 2

0.0

0.1

0.2

0.3

0.4

0 500 1000 1500 2000 2500 3000

Time / h

Cor

rosi

on ra

te /

(mm

/y)

"Dry

per

iod"

Sta

gnan

tContinuously exposed

Pre-oxidized

Figure 6 - Corrosion rate as a function of time, 50 wt% DEG, 1 wt% NaCl, 6 bar CO2, pH

6.5, 20 °C a) continuous flow, no precorrosion b) stagnant flow the first 10 days, c) "dry" period where the bulk liquid phase is drained from the system d) oxide covered specimens.

0

2

4

6

8

10

200 300 400 500 600 700

Tempering temperature / °C

Cor

rosi

on ra

te /

mm

/y

6.8 m/s4.1 m/s

0

20

40

60

80

100

200 300 400 500 600 700

Tempering temperature / °C

% s

urfa

ce c

over

ed b

y fil

m 6.8 m/s4.1 m/s

Figure 7 - Left: Corrosion rate as a function of tempering temperature for St 52 steel (C:0.15, Si:0.18, Mn:1.57,S:0.011,Cr:0.03, Ni:0.04, Cu:0.15) coupons. Values for specimens with mesa attacks are encircled. Right: Film coverage measured after exposure.

27 Appendix 2

80

85

90

95

100

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16

[Carbon] / %

Inhi

bito

r effi

cien

cy /

[%] A

M

E D

K

L

J

N

G

B

P

I

H

C

O

F

Figure 8a - Inhibitor efficiency of 40 ppm QUAT plus 0.5 ppm THIO after 150 h precorrosion plotted vs. carbon concentration. Filled symbols: F-P steels, open symbols: QT steels. Experimental conditions: 25 C, pH 5, 1 bar CO2, 3 wt% NaCl.

0

20

40

60

80

100

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

[Copper] / wt%

Inhi

bito

r effi

cien

cy /

%

[C] < 0.1 %[C] > 0.1 %

K O

A

C

BI

E

D

Figure 8b - Inhibitor efficiency of 20 ppm imidazoline based Inhibitor A) after 10 days

precorrosion vs. copper concentration. Experimental conditions: 25 °C, pH 5, 1 bar CO2, 3 wt% NaCl.