SciELO - Scientific Electronic Library Online

 
vol.34 número2Corrosão atmosférica do aço macio em atmosferas marinhasInfluência da cor de tintas de poliuretano na absorção e na dissipação de calor índice de autoresíndice de assuntosPesquisa de artigos
Home Pagelista alfabética de periódicos  

Serviços Personalizados

Journal

Artigo

Indicadores

Links relacionados

  • Não possue artigos similaresSimilares em SciELO

Compartilhar


Corrosão e Protecção de Materiais

versão On-line ISSN 2182-6587

Corros. Prot. Mater. vol.34 no.2 Lisboa dez. 2015

 

ARTIGO

Study of the influence of the water ph on the determination of corrosiveness of petroleum derivative on carbon steel

Estudo da influência do ph da água na determinação da corrosividade de derivados de petróleo ao aço-carbono

 

V. Y. Nagayassu*1, Z. Panossian2 and M. M. Seckler3

 1 Instituto de Pesquisas Tecnológicas do Estado de São Paulo S.A., Centro de Tecnologia em Metalurgia e Materiais, Laboratório de Corrosão e Proteção, Av. Professor Almeida Prado, 532, 05508-901, São Paulo, Brasil

2 Instituto de Pesquisas Tecnológicas do Estado de São Paulo S.A., Diretoria de Inovação e Universidade de São Paulo, Escola Politécnica, Departamento de Engenharia Metalúrgica e de Materiais, 05508-901, São Paulo, Brasil, e-mail: zep@ipt.br

3 Universidade de São Paulo, Escola Politécnica, Departamento de Engenharia Química, 05508-010, São Paulo, Brasil, e-mail: mseckler@abeq.org.br

* A quem a correspondência deve ser dirigida, e-mail: vann@ipt.br

 

ABSTRACT

Light petroleum derivatives consist of nonpolar hydrocarbons, which are aggressive to carbon steel due to the presence of water. In this study, we show both experimentally and theoretically that the water-phase pH during corrosion is controlled by the interaction of the aqueous phase with the atmospheric CO2 and with iron corrosion products in both solution and precipitates. During the corrosion process, the pH of the solution changes, regardless of its initial value, towards a neutral value. This behaviour explains why the corrosiveness of a petroleum derivative emulsion is not influenced by the initial water-phase pH.

Keywords: Carbon Steel, Fuel Corrosiveness, petroleum Derivate, Naphtha

 

RESUMO

Derivados claros de petróleo consistem de hidrocarbonetos não polares os quais são agressivos ao aço-carbono devido à presença de água. Neste trabalho, será mostrado experimentalmente e teoricamente que o pH da fase aquosa durante a corrosão é controlado pela interação da fase aquosa com o CO2 atmosférico e com os produtos de corrosão de ferro tanto em solução, quanto no precipitado. Durante o processo de corrosão, o pH da solução varia, independentemente do seu valor inicial, no sentido de um valor neutro. Este comportamento explica a razão da corrosividade de uma emulsão de derivado de petróleo não ser influenciada pelo pH inicial da fase aquosa.

Palavras-chave: Aço-carbono, Corrosividade, Derivado de petróleo, Nafta

 

1. INTRODUCTION

Light petroleum derivatives such as gasoline and naphtha consist mainly of nonpolar hydrocarbons, which are not aggressive to carbon steel in the absence of water (1,2). However, these derivatives are often mixed with water, which is originated during the refining processing and during storage in nonhermetic tanks (1). Water is slightly soluble in light petroleum derivatives, and thus it partly emulsifies. The amount of emulsified water depends, among other factors, on the degree of stirring. A large amount of water may be kept emulsified as long as a high degree of stirring is maintained. When stirring ceases, the water droplets coalesce and form an aqueous phase beneath an organic phase. In such conditions, light petroleum derivatives become very corrosive to carbon steel. Knowledge of the corrosiveness of such petroleum derivative water emulsions is important in relation to the physical integrity of carbon-steel tanks and pipelinStudy of the influence of the water ph on the determination of corrosiveness of petroleum derivative on carbon steeles used for their storage and transportation, respectively. Therefore, before petroleum derivates (gasoline and distillate fuels) are transported through pipelines, their corrosiveness is determined by means of the NACE standard TM-0172 (3). This test consists of visual evaluation of the corroded area of carbon steel specimens (coupons) after their immersion for 4 h, under stirring, in the tested petroleum derivatives previously mixed with 10 % distilled water of pH 7.0. The degree of corrosiveness of the derivates is given as the percentage of the specimen surface area that is corroded, as opposed to other methods that are based on the weight loss.

Today, the role of water in the corrosiveness of light petroleum derivatives is widely accepted. In view of the difficulty of preventing water intake, the corrosion is usually controlled in the petrochemical industry by adding corrosion inhibitors to the white petroleum derivatives. Thus, much research in this area focuses on the efficiency of corrosion inhibitors (a review on the subject is given in (4)).

An extensive literature search revealed that there are many fundamental studies related to crude petroleum-water mixtures, but studies related to petroleum derivative-water emulsions are rare. The interest in petroleum-water mixtures is probably due to the complexity of the formation water present in crude petroleum, which exhibits high concentrations of a great variety of dissolved salts and gases, such as chloride, CO2 and H2S (1, 5 - 8), while the water in light petroleum derivatives presents a lower concentration of corrosive impurities (1,9).

Kirkov (9) has mentioned that the solubility of oxygen and ionic species is very low in petroleum derivatives, and thus the expected corrosion rates of carbon steel are very low. He pointed out that the presence of moisture in petroleum derivatives increases the rate of carbon steel corrosion to a higher degree than that in which HCl affects the dissolution of metals in an aqueous solution. According to Kirkov, this fact indicates that the mechanism of metal corrosion in a light petroleum derivative is a more complex phenomenon than the electrochemical processes of metal dissolution in aqueous solutions.

The deleterious effect of moisture on the corrosiveness of petroleum derivatives was argued by Foroulis (1) as well as by groysman and Erdman (2), however, the role of oxygen was not. According to these authors, the solubility of oxygen in crude oil is very low, but it becomes considerably high in light refined products such as gasoline and naphtha, reaching values between 60 ppm and 70 ppm. These values are much higher than the solubility of oxygen in water which is almost 8 ppm. The oxygen solubility in different nonaqueous fluids was investigated by Hale (10), who confirmed the high solubility of oxygen in light hydrocarbons.

groysman and Erdman (2) concluded that the main cause of the corrosiveness of light petroleum products (naphtha and gasoline) is the presence of water and dissolved oxygen. They found that the corrosion mechanism in the naphtha-water or gasoline-water mixtures was similar to the mechanism of atmospheric corrosion under a wet film of electrolyte. In this case, naphtha and gasoline provided a reservoir of oxygen (instead of the atmosphere) and corrosive species such as chlorides, sulphates, nitrates and organic acids. In some instances, naphtha can also be considered as a reservoir of corrosive organic compounds, which are produced by its bacterial degradation (11,12).

From the above discussion, it appears that the corrosiveness of petroleum derivative emulsions has been generally associated with the presence, in the aqueous phase, of dissolved oxygen, of compounds transported from the organic phase and of organic compounds produced by bacterial activity. given the low concentrations of such corrosive impurities (1,9), it is likely that the composition of the aqueous solution is also influenced by its interaction with external systems other than the petroleum derivatives, such as the air and the metal surface with, for example, the atmospheric carbon dioxide and corrosion products such as iron hydroxides, respectively. These compounds participate in chemical reactions in the aqueous phase that may consume or produce hydroxyl ions, thereby affecting the water pH.

In brazil, the residual water found in pipelines is often of an acidic nature (pH of about 4). Therefore, the NACE method has been modified by the local oil industry to better simulate this condition: the distilled water added to the derivatives is previously acidified to a pH of 4.5 with HCl.

Surprisingly, Ferreira et al. (13) found experimentally that the pH does not affect the corrosiveness of such emulsions. We also came to the same results in a comprehensive study that took into account, in addition to the water pH, the water amount, temperature, stirring intensity and corrosion time (14). In the present work, we further investigate the effect of the water pH on the corrosion of carbon steel in the presence of petroleum derivative emulsions. We explain, both experimentally and theoretically, how the interactions of the water-phase with atmospheric carbon dioxide and with the corrosion products determine the solution pH in the course of the corrosion process.

 

2. EXPERIMENTAL

All the experiments were conducted using commercial grade naphtha and distilled water following the NACE TM-0172 (3) test method. Therefore, all the materials and equipment described in the test method were used, which included a cylindrical AISI 1020 carbon steel specimen with a conical base (81.0 mm x 12.7 mm), a grinding and polishing apparatus, a modified 400 mL berzelius-type glass beaker (Fig. 1a), an appropriate beaker cover, a rust-preventing oil bath and a thermometer.

 

 

The NACE test method (3) presents a step-by-step test procedure. A brief description of this procedure is as follows:

a. Specimen preparation: the specimens were cleaned with acetone, mounted in the chuck of the grinding and polishing apparatus and rotated at speeds ranging from 1,700 rpm to 1,800 rpm while a 100 grit J-weight silicon carbide abrasive cloth was applied. The polishing was done until a uniform, finely scratched surface free from longitudinal scratches was obtained.

b. testing procedure: 300 mL of the petroleum derivative to be tested was added into the beaker, and the beaker was placed in the oil bath, which was maintained at a temperature that kept the petroleum derivative at (38±1) oC. The cover, the specimen, the thermometer and the stirrer were adequately installed, and the petroleum derivative was submitted to stirring at (1,000±50) rpm for 30 min to ensure complete wetting of the steel specimen. After 30 min, distilled water (30 mL) was added. The stirring was maintained for a 3.5-hour period. After the completion of the test, the stirring was stopped and the specimen was removed. The specimen was washed with xylene followed by acetone.

c. Classification of the tested petroleum derivative corrosiveness: each specimen was inspected with a lighted glass of 2x magnification, and its corroded area was visually estimated.

Initially, the effects of the following variables on the corrosiveness of naphtha-water emulsions were studied: water-phase pH, water amount, temperature, stirring intensity, test period and carbon steel specimen geometry.

Simple immersion experiments in naturally aerated distilled water with different pH values were also conducted using the above-mentioned beaker (Fig. 1a). Additionally, an immersion experiment in nonaerated distilled water was conducted using an hermetic cell with a cover, which was provided with valves made of glass and polytetrafluoroethylene (pTFE), which were used to introduce ultrapure nitrogen gas to the tested water (Fig. 1b). In the interior of the cell, a test specimen holder made of polytetrafluoroethylene (PTFE) was inserted. The latter experiment was conducted using an AISI 1020 carbon steel plate (60 mm x 20 mm). In these experiments, the above-mentioned specimen preparation procedures were adopted.

Diluted hydrochloric acid and sodium hydroxide solutions were used to vary the water pH. All the reagents used were analytical grade.

The corrosion product composition was determined by X-ray diffraction using a Rigaku, model RINT 2000.

 

3. RESULTS

An experimental design with all considered variables was executed. Table 1 shows the selected variables and their levels. It was found that 216 experiments would be needed. To reduce the number of experiments, an asymmetric fractional factorial design of 33.23//36 was implemented. Thus, it was possible to evaluate the main effects and the interactions between two factors (independent variables) in an answer expressed as the percentage of a corroded area of carbon steel specimens (dependent variable). Table 2 shows the established design of the experiments. Each independent variable assumed all the levels considered, taking into account, as mentioned previously, combinations of each of the two selected factors.

 

 

 

The results obtained for the experiments shown in Table 2 were analysed, and, among other conclusions described elsewhere (4, 14), no influence of the water-phase pH on the corroded area was found. Similar results were found by Ferreira et al. (13). These results were unexpected because it is widely known that acidic media are more aggressive to carbon steel than neutral and alkaline solutions. To clarify this unexpected behaviour, the results related to water pH influence were carefully analysed. Thus, the water pH values after completion of the experiments conducted under the 36 described conditions were compared to the pH values before the experiments, as shown in Table 3.

 

 

Observation of Table 3 reveals that:

a) Regardless of the initial pH, the aqueous phase tended to assume a value close to neutral. b) In general, the final pH values for experiments initiated with acidified water were slightly lower than those initiated with neutral or alkaline water.

Three additional experiments were conducted, each of which utilised initial water pH values of 4.5, 7.0 and 10.0, for which the pH of the aqueous phase was measured every 10 minutes. The results are shown in Fig. 2.

Observation of Fig. 2 shows that:

 

 

a) For the three cases, the water pH varied towards a value close to neutral.

b) For an initial water pH of 4.5, a rapid increase in water pH took place initially until a value close to neutral was reached. Subsequently, the pH stabilised close to neutral.

c) For an initial water pH of 7.0, a fast initial increase in water pH took place. In this case, it occurred until a pH value of about 8.0 was reached. Subsequently, the pH decreased slowly towards values close to neutral.

d) For an initial water pH of 10.0, the pH dropped rapidly at first, and thereafter it decreased slowly towards values slightly higher than neutral.

Firstly, the issue was raised about the nature of the cathodic reaction: in the experiments, an intense stirring was applied, and thus the water added to the derivative became saturated in oxygen, leading to the supposition that oxygen reduction would be the predominant cathodic reaction. The only exception may be the experiment with initial water pH of 4.5. At this pH, hydrogen reduction also may occur. However, as can be seen in Fig. 2, in the beginning of this experiment, a deep increase of the pH occurred reaching values higher than 6.0. At this pH, the main cathodic reaction is also the oxygen reduction.

In order to verify this hypothesis, polarisation curves were experimentally determined for carbon steel immersed in water in stirring conditions similar to the experiments. The limiting current found for oxygen reduction was of the order of 1.10-4 Acm-2. Using this information and the electrochemical parameters (shown in Fig. 3) of the corrosion reactions of carbon steel, theoretical polarisation curves were built for the three pH values considered. These curves are shown in Fig. 3.

 

 

The electrochemical parameters used in Fig. 3 were (15):

Fig. 3 shows that, for water pH values of 7.0 and 10.0, the reaction of oxygen reduction determines the rate of carbon steel corrosion, whereas the influence of the hydrogen cation reduction is expected for the pH of 4.5. However, as the water pH rapidly approaches neutrality, the action of hydrogen reduction becomes negligible.

Based on the above results, it was concluded that the reaction of oxygen reduction is indeed the predominant cathodic reaction for water pH values of 4.5 to 10.

The following questions still remain:

a) Why does the water pH tend to reach a value close to neutral, regardless of the initial pH?

b) Is this variation of water pH related to some action of the petroleum derivatives?

To investigate these questions, three additional experiments were performed with distilled water, without a petroleum derivative addition, with water pH values of 4.5, 7.0 and 9.5, adopting a stirring speed of 1,050 rpm and a temperature of 25 oC. For each experiment, the water pH was measured every 10 minutes. Additionally, the pH was measured at the end of the experiments before and after boiling the aqueous solution.

The results are shown in Fig. 4 The pH variation with time was almost the same as in the case of the petroleum derivative-water emulsion presented previously (Fig. 2). Therefore, the variations in water pH are not related to the presence of the derivatives. Fig. 4 also shows that, in all cases, the water pH changed after boiling. The water pH being affected by the absorption of carbon dioxide from the atmosphere is a well-known fact.

 

 

The following question still needs an explanation: is the water pH variation towards neutral values somehow related to carbon steel? To answer this question, three additional experiments were performed, similar to the previous ones, but in the absence of the carbon steel specimen. The water pH was measured every 10 minutes. The results are shown in Fig. 5.

 

 

Fig. 5 shows that, for the initial water pH value of 9.5, the pH decreased with and without the carbon steel specimen, showing that this behaviour does not depend on the presence of carbon steel (compare with Fig. 2), unlike the experiment at a water pH of 4.5 and 7.0. For the latter, an initial increase followed by a slow decrease was observed in the presence of the carbon steel specimen (Fig. 2) and only a decrease of the pH was observed without the carbon steel specimen. The initial increase in the presence of carbon steel is a consequence of the alkalization resulted from the reduction of dissolved oxygen due to the corrosion process and the decrease for both experiments is due to carbon dioxide absorption.

In the case of pH 9.5, an initial alkalization in the presence of the carbon steel specimen is not observed because the ferrous ions generated by the corrosion reaction reacted with the hydroxyl ions generated by the oxygen reduction because the pH is higher than 9.2.

For a water pH of 4.5, no change in the pH was observed without the carbon steel specimen (Fig. 5) and an increase of the pH is observed in the presence of carbon steel specimen (Fig. 2).

It is convenient to mention that in all experiments described up to this point, the corrosion products adhering to the surface of the specimens as well as those in solution assumed a reddish colour, indicating the presence of ferric ions, which was actually expected, as all experiments were performed in open cells under intense stirring and therefore in aerated media. For confirmation, the corrosion products formed at the surface of one specimen, as well as those present in the aqueous solution, were analysed by X-ray diffractometry. The corrosion products were found to be composed of ferric oxides containing ferrous oxides, namely:

a) surface products: lepidocrocite (basic iron III oxide), magnetite (oxide of iron II and III) and maghemite (oxide of iron III);

b) aqueous phase product: lepidocrocite.

 

4. THEORY AND DISCUSSION

The theoretical analysis to elucidate the results obtained in all experiments was done by considering the following systems: pure water in contact with the atmosphere (aerated condition), pure water in contact with iron and with (aerated condition) and without (nonaerated condition) contact with the atmosphere and pure water in contact with iron and with a CO2-free atmosphere (aerated).

Variation of the pH of pure water in contact with the atmosphere (Aerated Condition)

The experiments shown in Fig. 5 were conducted using distilled water with three different pH values in contact with atmospheric air, i.e., in a natural aerated condition. In these cases, the observed pH variation can only be attributed to the absorption of atmospheric CO2.

Carbon dioxide is present in the atmosphere in concentrations of about 0.03 % (16). This gas dissolves in water and reacts with it, forming carbonic acid (H2CO3), which in turn dissociates to form the bicarbonate -ion HCO3 and the carbonate ion CO3-. Table 4 presents the main reactions occurring among the species present in the H2O-CO2 system as well as the equilibrium constants and concentrations (partial pressures).

 

 

Table 4 shows that the determination of the equilibrium pH requires the solution of a system of nonlinear equations. To this end, the method of intuitive approximation can be applied (18) in combination with simplifying hypotheses. The consistency of the solution is verified against the original equations. If the solution is unacceptable (because the hypotheses are not suitable), successive approximations are performed until an acceptable solution is found. The solution in each attempt is used as a guess for the next attempt. A more complete and accurate solution is obtained numerically. As an example, the equilibrium pH value of distilled water with a pH of 7 in contact with atmospheric air was calculated using the method of intuitive approximation. This calculation is shown in APPENDIX A.

The software MINTEQA2/pRODEFA2 (19) was used to calculate the equilibrium pH of the water in contact with atmospheric air for various initial water pH values (including pH of 7) to simulate the conditions found in the experiments of Fig. 5. The results are shown in Table 5.

 

 

It can be seen that the estimated values when considering CO2 absorption are very close to the experimental values. Therefore, one can conclude that the variations in water pH observed when distilled water is brought into contact with atmospheric air are due to CO2 absorption. The absorption is more important for higher water pH and is negligible for acidic water.

Variation of the pH for pure water in contact with Iron and with (Aerated Condition) and without (Nonaerated Condition) contact with a CO2-Free atmosphere

When an iron specimen is immersed in distilled water, even at pH 7.0, in contact with CO2-free atmospheric air (aerated condition), several reactions take place; therefore, the calculation of the final pH is more complex than in the previous discussion where iron was absent. Table 6 shows some of the possible reactions in this system, after iron corrosion, as well as their equilibrium constants. The dissolved species considered were Fe2+, Fe3+, H+ and OH-, and the precipitated solids taken into account were Fe(OH)2 and Fe(OH)3. Actually, many other compounds can be formed, such as soluble complexes (e.g., FeOH+) or solids such as magnetite (Fe3O4), hematite (Fe2O3), goethite (a-FeOOH), lepidocrocite (g-FeOOH) and ferrous oxide FeO. The simplification of adopting only Fe(OH)2 or Fe(OH)3 is justified based on the fact that the free energy of formation of the Fe3+ compounds (main constituents of the corrosion products) are of the same order of magnitude among each other, and similarly for the Fe2+ compounds (17, 20).

 

 

 

To better understand the variations in water pH observed in the experiments of Fig. 4, two simplified conditions were considered: the Fe/H2O system isolated from atmospheric air with a predominance of Fe2+ species and the same system in contact with CO2-free atmospheric air with a predominance of Fe3+ species. Thereafter, the observed pH variations will be explained.

 Fe/H2O system isolated from atmospheric air (water with pH 7.0)

The reactions responsible for the iron corrosion are the following:

From the equations above, one can conclude that corrosion promotes the alkalinisation of water, a well-established fact. These reactions proceed until the solubility limit of Fe(OH)2 is reached. At this point, precipitation of Fe(OH)2 takes place, i.e.,

Once formed, solid Fe(OH)2 remains in equilibrium with the soluble species that it originated from.

Using the method of successive approximations, it is possible to calculate the saturation pH for Fe(OH)2, i.e., the pH at which Fe(OH)2 precipitation starts (see APPENDIX B). A pH value of 9.2 was derived, which is the same as the literature value for the pH of a saturated Fe(OH)2 solution (20). During the corrosion of iron in water with pH 7.0, an increase of pH values is observed until the saturation point of Fe(OH)2, which occurs at pH 9.2. After this point, the water pH remains constant. Fig. 6 illustrates this fact by means of calculations with the software MINTEQA2/ pRODEFA2 (19).

 

 

If the initial water pH were lower than 7.0, the same behaviour would be observed, i.e., the corrosion of iron would cause an increase in the pH and in the concentration of Fe2+ up to the solubility limit of Fe(OH)2. The amount of Fe2+ ions necessary to reach saturation would be greater for smaller initial pH values.

If the initial pH is higher than 9.2, the solution rapidly becomes supersaturated with respect to solid Fe(OH)2 as soon as the corrosion starts. The pH might remain above 9.2 in a super saturation condition. However, after some time, Fe(OH)2 precipitation takes place, with a consequent pH decrease to the equilibrium value of 9.2.

As an example, the amount of Fe2+ necessary for the formation of the first Fe(OH)2 crystal in water, which has an initial pH of 9.5, is calculated below:

Fe/H2O system in contact with CO2-Free atmospheric air (water with pH 7.0)

Under these conditions, the reactions responsible for iron corrosion are

It is possible to calculate, by the method of intuitive approximation, the equilibrium pH of this system by assuming that after oxidation of both metallic iron and the total amount of generated Fe2+ ions, precipitation of Fe(OH)3 takes place immediately upon saturation of the solution (see Appendix c). Once formed, solid Fe(OH)3 remains in equilibrium with the soluble species that it originated from.

Because the solubility of Fe(OH)3 is very low (see Table 6), all Fe3+ and OH- ions produced by iron corrosion react with each other to form Fe(OH)3, leaving the solution almost free of these ions. Therefore, iron corrosion in these circumstances would not cause the pH to change. This assertion can be better understood from the partial and total equations given below:

A very important fact to note is the oxidation of Fe2+ into Fe3+ ions. In practice, this reaction is never complete, so a residual amount of Fe2+ ions remains in solution. However, this reaction is favoured in acidic media because it produces OH- ions. Therefore, during iron corrosion, the more acidic the solution, the more pronounced the oxidation of Fe2+ into Fe3+.

Fe/H2O System in contact with CO2-Free atmospheric air, with partial oxidation of Fe2+ into Fe3+ (water with pH 7.0)

The problem now proposed is very similar to the experiment shown in Fig. 4 after boiling. by application of the previously presented results to the Fe/H2O system in contact with CO2-free atmospheric air, the following reactions would take place:

a) Step 1: iron corrosion with generation of Fe2+ ions and consumption of H+ ions (or generation of OH- ions). This process leads to a gradual increase of the solution pH.

b) Step 2: oxidation of some of the Fe2+ ions formed in Step 1 into Fe3+ ions, followed by precipitation of Fe(OH)3. The formation of Fe3+ and the precipitation of Fe(OH)3 do not by themselves alter the solution pH.

c) Step 3: saturation of the solution with Fe2+ ions and precipitation of Fe(OH)2. From this point on, the solution pH remains at a value of 9.2.

Therefore, the final pH of the solution is governed by the amount of dissolved Fe2+ ions that have not yet been transformed into Fe3+.

In short, when an iron specimen is immersed in water in contact with CO2-free atmospheric air, the water pH might, for a finite period of time, reach a value between the initial value and the saturation pH 9.2 for Fe(OH)2. The more alkaline the solution, the closer the pH approaches the Fe(OH)2 saturation value; the more acidic the solution, the closer the pH remains at the initial value.

Variation of the pH for pure water in contact with iron and with CO2-containing atmospheric air

Finally, it is necessary to consider the presence of CO2 in the atmosphere. The system in question consists of an iron specimen immersed in pure water (pH 7.0) in contact with air, i.e., in contact with both O2 and CO2.

The first question that arises is whether the CO2 absorbed in the aqueous solution remains in dissolved form (HCO3 or CO32-) or precipitates as Fe2+ or Fe3+ iron compounds, such as FeCO3 (Kps = 3.13×10-11) (22).

For pure water (pH 7.0), the amount of CO32- present following atmospheric CO2 absorption is very small, of the order of 3.75×10-11 mol L-1 (see APPENDIX A). The corresponding amount of Fe2+ ions necessary for FeCO3 precipitation is very high, so it can be assumed that no precipitation occurs. pourbaix(20) described an experiment in which an iron specimen was immersed in an aerated solution saturated with sodium bicarbonate (pH 8.4). Under these conditions, iron remains passive due to the formation of iron oxides/hydroxides upon its surface, without the formation of carbonate-containing compounds. In addition, simulations performed for the conditions of the experiment shown in Fig. 4 with the software MINTEQA2/pRODEFA2 (19) revealed that iron carbonate does not form.

Therefore, one can assume that under the experimental conditions investigated here, all carbonate absorbed from the air remains in dissolved form as HCO3 and CO32- ions, the amount of the latter being negligible. Therefore, pH variations are consequences of CO2 absorption, iron corrosion and the precipitation of iron oxides/hydroxides.

Discussion of the experimental results

The experimental results previously presented in Fig. 4 and Fig. 5 are summarised in Table 7. This table shows that water with a pH of 4.5 in contact with air does not undergo a change in pH. This finding is expected because the only possible cause for pH variation is CO2 absorption from air, which is negligible under these acidic conditions.

 

 

However, when a steel specimen is immersed in this same water, the pH immediately starts to increase (Fig. 2) due to steel corrosion, as expected. However, the pH stabilises close to the neutral value because there are reactions that either produce or consume OH. The former consists of the cathodic reactions (reactions 2 and/or 4) and of the oxidation of Fe2+ into Fe3+ (reaction 5). The latter includes the precipitation of iron oxides/ hydroxides (reactions 3 and 6), which is favoured due to alkalinisation.

With neutral (pH 7.0) and alkaline (pH 9.5) water, the contact with air is sufficient to promote a decrease in pH, which is more pronounced for the most alkaline water of both. This pH decrease is a consequence of the absorption of atmospheric CO2, where the most alkaline water absorbs more CO2. The pH decrease persists in the presence of an iron specimen, even though the iron corrosion contributes to an increase of the pH. The role of CO2 is confirmed by the observed increase of water pH when CO2 is removed by boiling (Fig. 4).

For solutions (pH 4.5, 7.0 or 9.5) kept in an inert atmosphere, the final equilibrium pH is determined by the solubility product of ferrous oxide/ hydroxide, which is 9.2.

The factors that control the water-phase pH variation during corrosion of carbon steel in petroleum derivative emulsions are now understood: the oxidation of Fe into Fe2+ and of Fe2+ to Fe3+, the precipitation of iron oxides/ hydroxides and the absorption of atmospheric CO2.

Therefore, the interactions of the aqueous phase of the petroleum derivative emulsions with the atmospheric carbon dioxide and with the corrosion products determine the changes in pH during the corrosion processes. Understanding such pH variations may largely contribute to the establishment of corrosion mechanisms and kinetics.

 

5. CONCLUSIONS

The pH of the aqueous phase in petroleum derivative emulsions is controlled during corrosion by the interactions of water with the atmosphere and with corrosion products. In aerated conditions, such interactions are represented by the absorption of CO2 from the air and by the oxidation of iron, respectively. The latter involves the transportation of Fe2+ ions into the aqueous solution, their partial oxidation to Fe3+ and the subsequent precipitation of some of the Fe2+ and Fe3+ ions as iron oxides/hydroxides. The

chemical reactions and phase equilibria are such that, during the corrosion process, the pH of the solution changes, regardless of its initial value, towards a value close to neutral. The timescale for such pH changes is of the order of minutes. Therefore, the interactions of the aqueous phase with the air and corrosion products explain why the corrosiveness of a petroleum derivative emulsion is not influenced by the initial water-phase pH.

 

REFERENCES

(1) Z. A. Foroulis, Werkst. Korros., 33, 121-131 (1982).         [ Links ]

(2) A. groysman, N. Erdman, Corrosion, 56, 1266-1271 (2000).         [ Links ]

(3) NACE Standard TM-017-2001. (Determining Corrosive properties of Cargoes in petroleum product pipelines), NACE International, Houston (2001).

(4) V. Y. Nagayassu, Master Degree Thesis, University of São paulo, brazil (2006).

(5) g. A. Lee, g. A. Haines, (Corrosion problems of the petroleum Industry), 1st Ed., Staples printers Limited, Rochester, England (1960).

(6).J. I. bregman, (Corrosion Inhibitors), 1st Ed., Macmillan Company, N.Y., USA (1963)        [ Links ]

(7) p. Altoé, C. F. Moulin, S. L. Díaz and O. R. Mattos, Electrochim. Acta, 41, 1165-1172 (1996).         [ Links ]

(8) J. L. benitez, C. Martínez and R. Roldan, Oil Gas J., 100, 66-70 (2002).         [ Links ]

(9) p. Kirkov, Electrochim. Acta, 32, 921-926 (1987). 10) J. M. Hale, Orbisphere Document, Technical bulletin No. 6 (2002).

(11) A. Rajasekar, S. Maruthamuthu, N. Muthukumar, S. Mohanan, p. Subramanian and N. palaniswamy, Corros. Sci., 47, 257-271 (2005).

(12) A. Rajasekar, S. ponmariappan and S. Maruthamuthu, Curr. Microbiol., 53, 374-381 (2007).

(13) p. A. Ferreira, g. S. pimenta, E. M. O. paiva, R. F. brito e E. O. pinto, (préseleção de Inibidores para Dutos de produção com Aplicação de Técnica Eletroquímica e gravimétrica), COTEQ 2002, Salvador, brasil (2002),

(14) Z. panossian, V. Y. Nagayassu, A. A. g. bernal and g. S. pimenta, (Improvement of the NACE Test for Determination of the Corrosive properties of gasoline and Distillate Fuels), in Proceedings of Corrosion 2009, paper No 09578, March, Atlanta, USA (2009).

(15) M. Ohba, phD Thesis, University of São paulo, brazil (2003).

(16) Science is fun. Carbon dioxide. Available in: http://scifun.chem.wisc.edu/chemweek/CO2/CO2.html . Accessed on June 28 (2010).

(17) M. pourbaix, (Atlas of electrochemical equilibrium in aqueous solutions), 2nd Ed., NACE, Houston (1974).

(18) b. H. Mahan, (University Chemistry), 1st Ed., Addison-Wesley (1970).

(19) MINTEQA2/pRODEFA2, A geochemical Assessment Model for Environmental Systems, Computer Sciences Corporation, Assessment branch, AscI Corporation (1990).

(20) M. pourbaix, (Lectures on Electrochemical Corrosion), 1st Ed., CEbELCOR (1973).

(21) J. b. Russel, (general Chemistry), Vol. 1, CIp-brasil, Mcgraw-Hill (1992).

(22) Faculté technologique de chimie. Available in: http://www.ktf-split.hr/periodni/fr/abc/kpt.html . Accessed on June 28 (2010).

 

ACKNOWLEDGEMENTS

The financial support from CNPq - National Council for Scientific and Technological Development (brazil) is gratefully appreciated. The authors also thank Renata Angelon brunelli, gislaine Maria bragagnolo Nunes and Johny Hernandes de Oliveira for their assistance with experiments.

 

APPENDIX A

 

APPENDIX B

 

APPENDIX C

 

Creative Commons License Todo o conteúdo deste periódico, exceto onde está identificado, está licenciado sob uma Licença Creative Commons