Mild steel corrosion behavior in a coastal megacity relevant to China Pakistan economic corridor

Abstract

This 2-year study investigated mild steel corrosion in Karachi, Pakistan, comparing findings with other major coastal cities. Karachi plays a strategic role in China Pakistan economic corridor (CPEC) project. International Organization for Standardization and American Society for Testing and Materials standards were used to measure atmospheric corrosivity (C), corrosion rates, time of wetness, pollutants, and corrosion products along with analytical techniques. Corrosion rates classified three urban test sites as C3, three marine test sites as C5, and two urban and two industrial test sites as C4. The power-linear function was used to predict corrosion rates and corrosivity categories over 20 years. Long-term predictions showed medium C3 for urban and industrial sites and high C4 for marine sites. Mild steel might be the most effective material at marine test sites (average corrosion rates of 383–416 µm y⁻¹). Different quantities/morphologies of lepidocrocite, goethite, and magnetite were commonly present, and akaganeite was occasionally detected.

Introduction

Steel is among the most valuable economic materials used in almost every kind of material for civil manufacturing, construction, military, and defense applications. The mechanical quality and low price of carbon steels have made them one of the most widely used building materials for prolonged exposure to the outdoor environment. Carbon steel is an essential component in automobiles, bridges, and ships. Mild steel is a type of low-carbon steel with excellent ductility and machinability. Despite these virtues, mild steel is vulnerable to corrosion, particularly under acidic or saline conditions¹, if not adequately protected by coatings, compared to other ferrous alloys containing chromium, molybdenum, and cobalt². Thermodynamically, the spontaneous oxidation of native metals above hydrogen in the electrochemical series is inevitable. Not surprisingly, corrosion costs worldwide are estimated at USD 2.5 trillion, which is about 3.4% of the global gross domestic product³. A cost reduction of 15–35% can be realized if corrosion control practices are adopted³.

Detailed characterization and fundamental studies on corroding processes are needed to curb such losses worldwide. Given the ubiquity of mild steel, it is essential to understand its short- and long-term corrosion behavior, especially near commercially and strategically critical coastal areas, due to high salinity, humidity, and other favorable conditions in those parts of the world^4,5,6. Several studies have explored various aspects of the corrosion behavior of low-carbon steels in marine and atmospheric conditions^7,8,9,10; however, very few long terms studies exist, mainly in soil or marine environments^11,12,13. The ISO 9224 standard provides a power-linear function approach for predicting long-term corrosion rates and corrosivity categories after extended exposure for up to 20 years¹⁴.

Karachi (Pakistan) is one of the major global coastal megacities, with a very dense population that caters economically to more than 1.6 million people. It borders the Arabian sea. Karachi’s indigenous contribution to national revenue is around 25%¹⁵. It is an integral component of the ongoing China-Pakistan Economic Corridor (CPEC) project, the flagship of China’s “One Belt One Road” initiative¹⁶. High anthropogenic and large-scale industrial activities, which generate atmospheric pollution in Karachi and high humidity, provide an ideal environment for rusting and deterioration of ferrous materials. Unfortunately, despite being a megacity with heavy economic contributions and a home to steel mills, very few studies exist on the atmospheric corrosion behavior of steel in this city^17,18,19,20. In the last few years, Bano et al. have published a series of research focusing on weathering and rusting on various steels in Karachi due to their economic importance^17,18,19,20. These studies comprised 1-year data and monitored the atmospheric corrosivity of different test sites in Karachi. The corrosion products were characterized by physicochemical methods. As a result of this study, relevant corrosion maps were generated to guide the researchers. For metallic structures exposed to air, such maps greatly assist the engineers and material scientists in correctly choosing construction materials and devising oxidation protection strategies²¹.

This work is a continuation of the goal to fundamentally understand the 2-year corrosion behavior of mild steel in a coastal area like Karachi by characterizing the products and keeping the local weather information in view at ten different sites (see Table 1 for geographical coordinates). The observational results of the first year of the study are published elsewhere¹⁸. The second-year data combined with previous data is reported here with detailed analysis as per ISO 9224, enabling the prediction of long-term corrosion rates and corrosivity categories for a wide range of environments in Karachi. As per the ISO and ASTM standards, exposure tests were performed to assess atmospheric corrosion, corrosion products, pollutants including chloride, sulfur dioxide, and corrosion rates. The atmospheric oxidation degradation products of mild steel after 24 months are investigated using advanced analytical techniques such as scanning electron microscopy (SEM) combined with energy-dispersive X-ray spectroscopy (EDS), Fourier transform infrared (FTIR) spectroscopy, and X-ray diffraction (XRD) to assess corrosion damage imparted to mild steel in an economic hub of the country in South Asia by natural and anthropogenic causes.

Table 1 Corrosivity category for mild steel derived from atmospheric pollutants and time of wetness by ISO 9223 norm for a wide range of environments of Karachi coastal city during the second year of study from July 2019 to June 2020.

Results and discussion

According to ISO 9223²² and ISO 9225²³, standardized corrosivity estimation is based on information on levels of the dominating environmental parameters: the temperature-humidity complex, pollution with SO_2, and airborne chlorides. Measurements of these parameters are required for the purpose of corrosivity estimation. In this work, the corrosion behavior of mild steel was examined at ten different selected exposure test sites of a coastal megacity (Fig. 1), which is an integral component of the ongoing CPEC project.

Fig. 1

Location of the ten different selected exposure test sites in Karachi coastal city: T₁-T₅ (urban test sites), T₆-T₇ (industrial test sites), and T₈-T₁₀ (marine test sites).

Deposition rates of atmospheric pollutants and time of wetness

Figure 2 shows the deposition rates of atmospheric pollutants (chloride and sulfur dioxide) and time of wetness in a wide range of environments of Karachi during the second year of study from July 2019 to June 2020. The annual average chloride deposition rates for the urban test sites category (T₁–T₅) were 85.6 mg m⁻² d⁻¹. For the industrial test sites group (T₆, T₇), 97.5 mg m⁻² d⁻¹ was observed. At marine-type test sites (T₈–T₁₀), the annual average deposition was logged as 94.6 mg m⁻² d⁻¹. Based on these chloride deposition rates, all the exposure test sites are placed in the S₂ category for moderate chloride deposition (60 < S₂ ≤ 300) as per ISO 9223²². Data concerning chloride deposition show small changes, which means that there is no significant influence of the distance to the seashore on chloride deposition. With few exceptions, chloride deposition always decreases when the distance to the shoreline increases¹¹.

Fig. 2

Annual average deposition rates of atmospheric pollutants (chloride and sulfur dioxide) and time of wetness in a wide range of environments of Karachi coastal city during the second year of study from July 2019 to June 2020. T₁–T₅ refer to urban test sites, T₆–T₇ refer to industrial test sites, and T₈–T₁₀ refer to marine test sites.

It is unusual to note that marine sites do not show higher chloride deposition than urban and industrial sites. This could be due to the climatology of Karachi. In addition to being a coastal city, there are two main rivers, namely the Liyari and Malir Rivers, and many streams, including Gujjar, Gogni, Thado, and Orangi, which flow through Karachi and drain into the Arabian Sea. These rivers and streams carry untreated wastewater of both domestic and industrial origins. Several industries lie in the vicinity of Karachi City, including shipping, railroad yards, automobile assembly plants, textiles, oil refineries, pharmaceuticals, cement factories, steel mills, tanneries, and several light industries¹⁸. In addition, many industries burn chlorine-containing coal to produce electricity. Thus, the burning of chlorine-containing coals and the emission of chlorine from industrial processes could be the reason for the higher chloride deposition at almost all test sites in the city; hence, airborne salt from the sea is not the only source of chloride.

These results of the current study for marine test sites are relevant to the marine test sites of India as both share the same Arabian Sea. The chloride burden of marine test sites for Karachi, Pakistan, was higher than that reported for Mangalore, India (Table 2)²¹. Marine sites near the Bay of Bengal, such as Cuddalore and Nagapattinam, also have a lower chloride burden (Table 2)²¹. It was also found that the marine sites of Karachi have more chloride burden than Colombia, Abu Dhabi, Saudi Arabia (Jubail, Khafji), China, Chile (Arica, Antofagasta, Huasco), and Venezuela (Table 2)^{8,24,25,26,27,28}. Results also indicated that the atmosphere of Karachi’s urban sites (T₁–T₅) and industrial sites (T₆ and T₇) contained a higher amount of chloride as compared to these sites of other countries (Table 2)^8,21,26,29.

Table 2 Comparison of deposition rates of atmospheric pollutants (Cl⁻ and SO₂) measured in the subject study with previous studies reported in the literature.

The annual average deposition rates of sulfur dioxide for the urban test sites category (T₁–T₅) and the industrial test sites group (T₆, T₇) were recorded as 155.4 mg m⁻² d⁻¹ and 186 mg m⁻² d⁻¹, respectively. At marine-type test sites (T₈–T₁₀), the annual average deposition rate of sulfur dioxide was logged as 118 mg m⁻² d⁻¹ (Fig. 2). The corrosion of metals increases considerably due to sulfur dioxide gases in the atmosphere²⁵. Burning sulfur-containing fuels, petrochemical, and industrial activities are primary anthropogenic sources of sulfur dioxide gas emissions in the city. In addition, traffic and wind patterns of a place dramatically influence the atmospheric SO₂ concentrations^11,25,30. ISO 9223 standard establishes four categories according to the sulfur dioxide deposition rate (P) expressed in mg/m²d; P₀ (P_d ≤ 10), P₁ (10 < P_d ≤ 35), P₂ (35 < P_d ≤ 80), and P₃ (80 < P_d ≤ 200)²².

Results of the present study have revealed that sulfur dioxide was present in high concentrations in the atmosphere of Karachi. All urban (T₁–T₅), industrial (T₆, T₇), and two marine test sites (T₈, T₁₀) were assigned the P₃/P₃-plus category except one marine site (T₉) that was allocated as the P₂ category. Data indicated that an urban (T₂) and an industrial (T₆) test site classification was higher than the P₃ category, i.e., P₃-plus. This may be because test site T₂ is located in Saddar, the central business hub in Karachi. It is a densely populated area with a very high traffic influx.

Similarly, test site T₆ is located in an industrial area with intense anthropogenic activities. Moreover, the marine test site (T₉) is very close to the creeks and mangroves of the Arabian Sea, which may be a reason for the moderate atmospheric level of SO₂ (P₂) found at this site. In addition, it was found that SO₂ deposition rates at marine test sites (T₈ to T₁₀) also agree with the previous research reported for marine test sites of Karachi²⁰. These results also indicated that Karachi’s urban, industrial, and marine atmospheric test sites were contaminated with a higher amount of SO₂, which is more corrosive than in many other countries (Table 2)^{8,21,25,26,27,28,29}.

From July 2019 to June 2020 (Fig. 2), the annual average TOW for the urban test sites category (T₁–T₅) and the industrial test sites group (T₆, T₇) have been recorded as 1922 h a⁻¹ and 2041 h a⁻¹, respectively. At marine-type test sites, the annual average TOW was logged as 3305.28 h a⁻¹. A comparison of the results indicated that the urban and industrial test sites fall in the τ₃ TOW category (250 < τ₃ ≤ 2500). In contrast, the τ₄ category was allocated to the marine test sites as per ISO 9223 (2500 < τ₄ ≤ 5500)²². The higher value of TOW observed for the marine test sites may be due to their closeness to the Arabian Sea, which is consistent with the findings of other researchers²⁸. An increase in TOW for marine sites could mean chloride deposition occurs in liquid form because chloride aerosols can be deposited as a saline solution on the surface of metals when humidity increases. Previously¹⁸, we applied multivariate regressions on first-year data to study the influence of monthly chloride (Cl⁻) and time of wetness (TOW) on the monthly corrosion rate of mild steel for ten different sites in Karachi city. The proposed model indicates the influence of chloride ions interacting with the TOW and is responsible for changes in the corrosion rate¹⁸. The influence of SO₂ and time of wetness (TOW) was also studied. During data processing, it was found that the p value of the independent variables belonging to SO₂ × TOW was ≥0.10, which means this term is not statistically significant at the 90% or higher confidence level¹⁸. Thus, no significant statistical effect was found for SO₂ and TOW. The deposition of SO₂ includes almost all sulfur compounds like SO₂, SO₃, sulfates, and H₂S. It is possible that the actual presence of the two most aggressive gases could change for the different sites, or the primary deposition is constituted by sulfate ions coming from the sea. Under these conditions, it is possible that the actual effect of SO₂ is not determined as statistically significant. TOW changes are relatively small for a territory like a city, so it is not easy to find a clear effect. Significant changes in TOW occur between different climates.

Appearance of corroded mild steel

The physical appearance of rust products is most intrinsic, and it depends on the composition and type of material exposed, the corrosivity of the atmospheric test station, and the time of exposure. Usually, a highly aggressive environment allows the passage of corrosive species and produces non-protective, loosely adherent rust on surfaces with cracks that can easily detach from the surface^31,32. During the second year of study, from July 2019 to June 2020, the surface features of exposed specimens were studied to understand the nature and intensity of corrosion. A comparison of the visual results has shown that the corrosion product that appeared on the mild steel test specimens was of assorted types for all the test sites (Fig. 3). Cracking, exfoliations, and bigger flakes were noticed on all test specimens repossessed from the marine test sites (T₈ to T₁₀) and have shown a considerable loss of thickness with the depositions of massive rust particles. These bigger flakes of dark brown colored rust were detached at these test sites due to more influential wind movement at marine sites. Highly penetrating pits were scattered on the surface of all test coupons retrieved from marine sites, which increased in size with time, particularly having a more significant size at the T₈ station. However, smaller and shallower pits and brown, powdery corrosion product grains were observed on the surface of test coupons exposed at all urban and industrial test sites (Fig. 3). These observations indicate that the marine test sites in Karachi are more aggressive than urban and industrial test sites due to the significant difference in TOW. The present study agreed well with the findings of other researchers, who found heterogeneous exfoliations and larger flakes formed at more aggressive test sites^31,32.

Fig. 3

The appearance of mild steel test coupons after 24-months of exposure in a wide range of environments of Karachi coastal city: T₁–T₅ (urban test sites), T₆–T₇ (industrial test sites), and T₈–T₁₀ (marine test sites).

Corrosion rates and kinetic parameters

Minimum, maximum, and average corrosion rates of mild steel and category of all test sites (T₁–T₁₀) during the second year of study from July 2019 to June 2020 are tabulated in Table 3. The lowest, highest, and average corrosion rates for the urban test sites category (T₁–T₅) have been recorded as 33 µm y⁻¹, 62 µm y⁻¹, and 46 µm y⁻¹, respectively. For the industrial test sites group (T₆, T₇), corrosion rates were the lowest at 44 µm y⁻¹, the highest at 62 µm y⁻¹, and the annual average of 51 µm y⁻¹. At marine-type test sites (T₈ to T₁₀), the lowest corrosion rate was recorded as 69 µm y⁻¹, the highest corrosion rate was founded as 98 µm y⁻¹, and the annual average corrosion rate was logged as 83 µm y⁻¹.

Table 3 Comparison of corrosion rates of mild steel exposed in a wide range of environments of Karachi coastal city during the second year of study from July 2019 to June 2020.

ISO 9223 standard establishes five corrosivity categories (C) according to the corrosion rates (Cr) expressed in µm y⁻¹; C1 (Cr < 1.3), C2 (1.3 < Cr ≤ 25), C3 (25 < Cr ≤ 50), C4 (50 < Cr ≤ 80), C5 (80 < Cr ≤ 200)²². Based on corrosion rates (Table 3), the corrosivity category C3 was assigned to three urban test sites (T₁, T₃, and T₅). On the other hand, two urban (T₂, T₄) and two industrial test sites (T_7, T₆) were categorized as C4, whereas the C5 category was allocated to all three marine test sites (T₈–T₁₀). A comparison of the results indicated that the corrosivity category established by measuring corrosion rates of carbon steel for the two urban (T₂ and T₄), two industrial (T₇ and T₆), and two marine test sites (T_8, T₁₀) are in accordance with the corrosivity category derived from pollutants (chloride and sulfur dioxide) level and TOW. However, the same relation was not proved suitable for the three urban test sites (T₁, T₃, and T₅) and a marine test site (T₉) (Tables 1 and 3)^9,28.

A literature study showed that these marine test sites (T₈–T₁₀) are more corrosive than the marine test sites Cartagena, Colombia (46 µm y⁻¹ > 26.52 µm y⁻¹), Khobar (46 µm y⁻¹ > 42.79 µm y⁻¹), Jubail (46 µm y⁻¹ > 36.69 µm y⁻¹), Khafji (46 µm y⁻¹ > 36.82 µm y⁻¹), Saudi Arabia, Punto Fijo (46 µm y⁻¹ > 21.92 µm y⁻¹), and Barcelona (46 µm y⁻¹ > 29.3 µm y⁻¹) Venezuela^8,25,28. Similar findings were noticed for the industrial (T₆, T₇) and urban (T₁–T₅) test sites of Karachi compared to other industrial and urban sites worldwide^21,25,33. Figure 4 shows kinetic parameters obtained for the mild steel exposed in a wide range of environments in Karachi during the second year of study from July 2019 to June 2020. For the urban test sites category (T₁–T₅), the value of “n” ranged between 0.21 and 0.39 (Fig. 4a). For both industrial test sites (T₆–T₇), the value of “n” was 0.29; for marine-type test sites (T₈–T₁₀), the value of n ranged between 0.33 and 0.48 (Fig. 4b). The value of the correlation coefficient (R²) for the urban test sites category (T₁–T₅) ranged between 0.91 and 0.98 (Fig. 4a), for both industrial test sites (T₆–T₇), it was 0.91, and for marine type test sites (T₈–T₁₀) it was ranged between 0.95 and 0.99 (Fig. 4b). A linear bi-logarithmic relationship between mass losses and the exposure time was observed for all test sites. This relationship may be used for forecasting the mild steel corrosion rates in the atmosphere of Karachi. The high value of R² (0.91–0.99) indicated a good correlation between the data.

Fig. 4: Log–log graph of mass loss versus exposure time for the mild steel exposed in a wide range of environments of Karachi coastal city during the second year of study from July 2019 to June 2020.

a Kinetic parameters ‘n’ and ‘R²’ for T₁–T₅. b Kinetic parameters ‘n’ and ‘R²’ for T₆–T₁₀. Here, T₁–T₅ refer to urban test sites, T₆–T₇ refer to industrial test sites, and T₈–T₁₀ refer to marine test sites.

From July 2018 to June 2019¹⁸, for all marine test sites (T₈–T₁₀), the value of n was greater than 0.5, which confirmed that the diffusion of active corrosive species was accelerated due to the detachment of the non-protective corrosion products by dissolution, flaking, and cracking^21,25. For all of the test sites between July 2019 and June 2020, the value of “n” was less than 0.5, which may be associated with the fact that mild steel corrosion is controlled by diffusion, and corrosion products are rather protective, which remain intact on the surface and prevent further corrosion^21,25 (Supplementary Table 1). Hence, a good correlation was observed between the visual examination results and kinetic study results. These findings were consistent with the previous studies^21,25. Comparasion of the maximum corrosion rates during the 1st year and 2nd year of study indicated that it was decreased with the passage of time²⁵ (Supplementary Table 2).

Long-term corrosion rates and corrosivity categories as per ISO 9224

The power-linear function used for predicting long-term corrosion rates of mild steel and corrosivity categories for a wide range of environments in Karachi¹⁴ is presented in Fig. 5a, b. The long-term prediction is medium C3 for the urban and industrial test sites (average predicted corrosion rates 225–249 µm y⁻¹), whereas high C4 for the marine test sites for 20 years. Thus, based on the corrosivity category, mild steel could be the most effective material at marine test sites (average predicted corrosion rates 383–416 µm y⁻¹). However, it is possible that corrosion products layering on the surface may cause a decrease in corrosion rate¹¹.

Fig. 5: Prediction of long-term mild steel corrosion rates (D) after extended exposure of up to 20 years calculated with Eq. (3) in a wide range of environments of Karachi coastal city.

a For urban test sites (T₁–T₅) b for the industrial (T₆–T₇) and marine test sites (T₈–T₁₀).

Corrosion product’s morphology and compositional analysis using SEM/EDS

SEM is successfully used to analyze the surface morphology of the corrosion products at different exposure sites^{8,31,34,35,36}. The general consensus is that rust comprises ferric oxyhydroxides and hydroxides, such as lepidocrocite, goethite, akaganeite, and magnetite. These products could coexist partly as crystalline and partly as amorphous substances and transform into each other in conducive conditions^31,32. A comparison of the results obtained after 24 months indicated that lepidocrocite, goethite, and magnetite were present in the corrosion products of all the test sites (Figs. 6, 7, and Table 4) (supplementary Fig. 1). It is well documented that lepidocrocite is considered harmful rust. Its porous morphology allows the penetration of the electrolyte and corrosive species, particularly SO₂ and NO_2, due to which the dissolution of lepidocrocite occurs. As a result, the concentration of ferric ions increases, which causes further oxidation of iron because of its oxidation ability. In the transformation of lepidocrocite to goethite, the role of chloride ions is vital³⁷. Magnetite formation occurs near the surface of a metal when lepidocrocite reacts with the Fe²⁺ ions produced by the oxidation of iron³¹. The findings of the present study are consistent with the results reported by other researchers^8,9,31,38.

Fig. 6: SEM images and corresponding EDS of the mild steel after 24-months of exposure at the urban test sites (T₁–T₅) of Karachi coastal city.

a, b T₁, c, d T₂, e, f T₃, g, h T₄, and i, j T₅. The corresponding EDS is shown next to the SEM figures.

Fig. 7: SEM images and corresponding EDS of the mild steel after 24-months of exposure at the industrial (T₆–T₇) and marine test sites (T₈–T₁₀) of Karachi coastal city.

k, l T₆, m, n T₇, o, p T₈, q, r T₉, and s, t T₁₀. The corresponding EDS is shown next to the SEM figures.

Table 4 Comparison of the SEM results of corrosion products formed on mild steel after 24 months of exposure in a wide range of environments of Karachi coastal city.

EDS analysis indicated that Si, Al, and Ca were present in the corrosion products of all the exposure test sites after 24 months of exposure (Table 5). Most likely, they exist in the outer part of the rust layer, while they both may be in the rust layer in the form of dust and impurities, which had little influence on the corrosion³⁹. A considerable amount of sulfur was noticed at most of the test sites. The primary sources of sulfur are vehicular traffic and the burning of biomass^11,25,30. Almost all test site specimens of mild steel showed traces increase in the mass% of Mg and a relatively modest presence of Cl. It has been reported that the accumulated Mg in the rust layer can replace some ferrous ions in the intermediates (γ-Fe.OH.OH)³⁹, although we did not see strong evidence in the characterization studies. The overall percentage of elemental Fe was decreased during the whole exposure period due to its consumption in the corrosion process.

Table 5 Elemental composition of the corrosion products formed on mild steel after 24-months of exposure in a wide range of environments of Karachi coastal city as determined by EDS analysis.

Identification of corrosion products through FTIR spectroscopy

A comparison of the FTIR spectroscopy results of corrosion products formed on mild steel after 24-months of exposure in a wide range of environments in Karachi is presented in Fig. 8 and Table 6. Results indicated that lepidocrocite, goethite, and magnetite were present in the corrosion products of almost all the test sites. Corrosion products found at all these test sites correspond well with the common corrosion products reported for mild steel in literature^8,9,25.

Fig. 8

Infrared spectra of the corrosion products formed on mild steel after 24-months of exposure in a wide range of environments of Karachi coastal city (a) T₁ (b) T₂ (c) T₃ (d) T₄ (e) T₅ (f) T₆ (g) T₇ (h) T₈ (i) T₉, and (j) T₁₀. T₁–T₅ refer to urban test sites, T₆–T₇ refer to industrial test sites, and T₈–T₁₀ refer to marine test sites.

Table 6 Comparison of the FTIR spectroscopy results of corrosion products formed on mild steel after 24-months of exposure in a wide range of environments of Karachi coastal city.

Detection of crystalline corrosion products by XRD

XRD confirmed the presence of lepidocrocite, goethite, and magnetite phases (associated 2θ reflections labeled in Fig. 9)^26,40,41 in the corrosion products formed on mild steel after 24 months of exposure at almost all the test sites of Karachi. In addition, akaganeite was also spotted by XRD at two test sites. A literature survey has shown that the composition of corrosion products could vary, depending mainly on the exposure conditions, identification techniques, and data interpretation^{8,31,32,41,42}. Corrosion products formed on mild steel exposed to different atmospheres greatly influence the corrosion rate. It has been reported that an increase in the relative magnetite content can be directly related to higher corrosion rates in long-term atmospheric corrosion⁴². The presence of akaganeite could also accelerate corrosion in Karachi city^40,42.

Fig. 9

X-Ray diffractograms of the corrosion products formed on mild steel after 24 months of exposure in a wide range of environments of Karachi coastal city (a) T₁ (b) T₂ (c) T₃ (d) T₄ (e) T₅ (f) T₆ (g) T₇ (h) T₈ (i) T₉, and (j) T₁₀. T₁–T₅ refer to urban test sites, T₆–T₇ refer to industrial test sites, and T₈–T₁₀ refer to marine test sites.

This work has provided the long-term corrosion behavior of mild steel at ten different selected exposure test sites of a coastal metropolis (Karachi, Pakistan), an important city of the China-Pakistan Economic Corridor, by applying ASTM & ISO standards in conjunction with spectroscopic techniques. A comparison of the corrosion rates of mild steel measured over the course of 24 months at different test sites in Karachi reveals that the metal infrastructure of Karachi is relatively vulnerable to atmospheric corrosion. Three urban test locations were assigned the corrosivity category C3 (medium) based on the corrosion rates of mild steel. Two urban and two industrial test sites were categorized as C4 (high), whereas the C5 (very high) category was allocated to all three marine test sites. A comparison of the results indicated that the corrosivity category established by measuring corrosion rates of carbon steel for the two urban (T₂ and T₄), two industrial (T₇ and T₆), and two marine test sites (T₈ and T₁₀) are in accordance with the corrosivity category derived from pollutants (chloride and sulfur dioxide) level and TOW. However, the same relation was not proved right for the three urban test sites (T₁, T₃, and T₅) and a marine test site (T₉). According to the power-linear model, the long-term prediction for urban and industrial test sites is a medium C3 and high C4 for marine test sites for 20 years. As a result, mild steel is expected to be the most effective material for marine test sites based on the corrosion category (average corrosion rates predicted to be 383–416 µm y⁻¹). Compared to global test sites, Karachi coastal city’s corrosivity and corrosion extent of mild steel were greater than those of industrial, urban, and marine test sites. There was no significant evidence about the effect of temperature on mild steel corrosion behavior. This is because temperature changes are almost negligible in a relatively small space like a city. However, the influence of temperature is observed in large spaces, particularly when different climates are compared.

At almost all test sites, SEM/EDS, FTIR, and XRD analysis have indicated the presence of lepidocrocite, goethite, and magnetite. In addition, akaganeite was spotted by XRD at two test sites. Thus the presence of akaganeite could accelerate corrosion in Karachi city.

Methods

Material selection and exposure tests

Commercial-grade mild steel test coupons were purchased from a local market in Karachi, Pakistan. Optical emission spectrometry (Bruker Q4 TASMAN series 2) was used to determine the composition of mild steel. The wt/wt% of different elements was C: 0.04%, Cr: 0.01%, Mn: 0.23%, Al: 0.03%, Si: 0.01%, Cu: 0.01%, S: 0.003%, P: 0.014% and Fe was present as remainder. Test coupons were machined into 103 × 152 × 3 mm dimensions, with smoothened side edges, and the burs were also removed. The surface of the test coupons was prepared as per ASTM G50-20 “Standard Practice for Conducting Atmospheric Corrosion Tests on Metals” protocol⁴³. Subsequently, test coupons were labeled, sealed into air-tight plastic bags, and transported to the selected atmospheric corrosion test sites. The weight of each test coupon was recorded before and after exposure testing to provide a baseline for comparison with the post-line test results.

Selection of test sites and exposure testing

In this study, the long-term behavior of mild steel was investigated in a wide range of environments. For this purpose, ten different test sites, including urban, industrial, and marine test sites, were selected all over Karachi. These test sites were labeled as T₁ to T₁₀. The location and physical features of these sites are depicted in Fig. 1 and tabulated in Table 1, respectively. Exposure racks were prepared following the ASTM standard practice G50-20⁴³. Test coupons were exposed at these test sites by mounting them on ceramic separators of the wooden racks at a 45° angle facing toward the sea for marine sites while facing south for other sites.

Deposition rates of atmospheric pollutants and time of wetness

During the second year of the study period, from July 2019 to June 2020, the deposition rates of atmospheric pollutants chloride and sulfur dioxide at the ten different selected exposure test sites were determined as per the ISO 9225:2012 norm, “Corrosion of metals and alloys—Corrosivity of atmospheres-Measurement of environmental parameters affecting corrosivity of atmospheres²³. The deposition rate of chloride ions was determined by the wet candle method, whereas the sulfation plates were used for sulfur dioxide. The results are reported in mg m⁻² d⁻¹. The detail of the procedures is already reported [17–20]. In order to determine the TOW, the Easylog EL-USB-2-LCD RH/Temp/ data loggers were installed at the ten selected exposure test sites. Daily temperature and relative humidity were recorded, and the time of wetness was calculated in h a^{−1 44}.

Appearance, corrosion rate, and kinetic parameters of mild steel

Test coupons were recovered after a specific time interval, and surface appearance was examined. The mass loss of each test coupon was recorded according to the ASTM G50-20 procedure⁴³. Cleaning was performed by ASTM standard practice G1-03: 2017, Standard practice for preparing, cleaning, and evaluating corrosion test specimens⁴⁵. First, light brushing under flowing water removed loosely adherent rust layers. Next, a mixture of hydrochloric acid and hexamethylenetetramine (C₆H₁₂N₄) was used to remove corrosion products from the test coupons at room temperature. Finally, the coupons were rinsed with de-ionized water and acetone and dried with hot air. After cleaning, the loss in mass was recorded by weighing residual coupons, and the temporal trend of mass loss was plotted for the entire exposure period. The corrosion rate was calculated in μm y^{−1 45}.

$${{{\mathrm{C}}}} = {{{\mathrm{K}}}}_{{{\mathrm{c}}}}\Delta {{{\mathrm{W}}}}/{{{\mathrm{AtD}}}}$$

(1)

where, C is the corrosion rate in μm y⁻¹, K_C has a constant value of 87,600, ΔW is the mass loss in mg, A is the area in cm², D is the density of steel (7.85 g cm⁻³), and t is the time of exposure in hours.

Corrosion kinetic parameter n and correlation coefficient (R²) were also determined²¹.

$$\log {{{\mathrm{C}}}} = \log {{{\mathrm{K}}}} + {{{\mathrm{nlog}}}}\,{{{\mathrm{t}}}}$$

(2)

Where ‘K’ is the intercept, n is the slope of the log–log plot, and t is the exposure time in hours.

Long-term corrosion rates and corrosivity categories as per ISO 9224

As per ISO 9224¹⁴, the prediction of long-term corrosion rates and corrosivity categories for a wide range of environments in Karachi was carried out by using the following relation:

$${{{\mathrm{D}}}} = {{{\mathrm{r}}}}_{{{{\mathrm{corr}}}}}\,{{{\mathrm{t}}}}^{{{\mathrm{b}}}}$$

(3)

where, “D“ is the total attack expressed as penetration depth in µm y⁻¹, “r_corr” is the corrosion rate experienced in the first year, expressed in µm y⁻¹, “t” is the exposure time, expressed in years, and b is the time exponent values for predicting and estimating corrosion attack which is 0.523 for mild steel.

Morphology of corrosion products and compositional analysis

The characterization of corrosion products formed on mild steel after 24 months of exposure was carried out by scanning electron microscopy (SEM)/Energy dispersive X-ray spectroscopy (EDS), Fourier transform infrared (FTIR) spectroscopy, and X-ray diffraction (XRD).

Scanning electron microscopy/ Energy dispersive X-Ray spectroscopy

In this study, the surface morphology of corrosion products formed after 24 months of exposure was assessed by SEM. Small pieces of 1 × 1 in. from exposed mild steel test specimens were cut, and the gold coating was applied using a gold coater (JEOL JFC). Images of test specimens were recorded, and the elemental composition of the corrosion products was obtained by using a JEOL 6380A microscope equipped with a secondary electron detector and JEOL EX-54175iMU EDS detector under the operating vacuum of 10⁻⁷ and 15 kV voltage.

Fourier transform infrared spectroscopy

The KBr pellet method was employed to measure the infrared spectra of corrosion products. For this purpose, the powdery corrosion product was ground with dried KBr salt using a mortar and pestle to make a disc. Shimadzu IR Prestige 21 Fourier transform infrared spectrophotometer running with Omnic software, in the range of 4000-400 cm⁻¹, was used to obtain the spectra. The spectrophotometer was operated in transmission mode, and a resolution of 4 cm⁻¹ and ten scans was found to be suitable.

X-Ray diffraction

The corrosion products were scraped from the surfaces of the exposed mild steel and test specimens and ground to make a fine powder using mortar and pestle. Bourevestnik Dron-8- multifunctional X-ray diffractometer equipped with a 2.5BSW27-Cu X-ray tube was used to obtain diffractograms. The voltage applied was 40 kV with a filament current of 30 mA. XRD beam incidence angles 2θ, range 10^o–90^o with a step width of 0.03° and a counting time of 5 s/step were used. Rietveld method for quantifying crystalline rust phases from the XRD pattern was employed⁴⁶.