A review of internal corrosion mechanism and experimental study for pipelines based on multiphase flow
-
Hao Zhang
Hao Zhang is a PhD candidate in vehicle operation engineering at Beijing Jiaotong University (BJU). He graduated from BJU in 2013 with a bachelor’s degree in mechanical design and automation. His research area focuses on the internal corrosion of pipelines in the petroleum industry.
and
Hui-qing Lan
Hui-qing Lan is a professor and a PhD supervisor at BJU. She completed her postdoctoral work at the University of Tokyo. She holds a PhD in engineering (2002) from the China University of Petroleum (Beijing) and is a member of the Petroleum Storage and Transportation Committee of the China Petroleum Society.
Abstract
The internal corrosion of pipelines in the petroleum industry is highly risky, and induced pipeline cracking may give rise to potential injury to personnel and environmental issues. The oil-water two-phase flow and the oil-gas-water three-phase flow are often observed in gathering and transportation pipelines. It is generally accepted that corrosion is induced by the presence of water, although it is a complex hydrodynamic process in which the material is removed from the pipeline due to physicochemical reactions. Hence, it is necessary to determine the key parameters that dominate the corrosion phenomena and how they can be modeled. As the water phase that wets the steel surface determines the initiation of corrosion, several aspects are widely discussed here, such as corrosive medium, phase inversion, water-wetting behavior, the entrainment of water, and the wettability of steel, to explain the corrosion mechanism of multiphase flow and correlation with the corrosion behavior. Of course, empirical and mechanistic models for corrosion prediction in pipelines are discussed. Also, the mostly applied techniques of identifying flow patterns and attaining related parameters in experiments for the evaluation of the corrosiveness of oil-brine mixtures are introduced. Further studies must be undertaken to expand the knowledge of corrosion and find applicable models for corrosion damage prediction and prevention.
1 Introduction
The internal corrosion of pipelines in the petroleum industry is highly risky, and induced pipeline cracking may give rise to potential injury to personnel and environmental issues, especially when the produced water in crude oil is highly corrosive with corrosive media, such as corrosive gas, microorganisms, and salts (Nesic & Sun, 2010). Once the steel surface is wetted by water, local corrosion will thin the pipe wall (Schmitt & Stradmann, 1998; Larsen, 2013). Hence, it is generally accepted that that the corrosion is induced by the presence of water, although it is a complex hydrodynamic process in which the material is removed from the pipeline due to physicochemical reactions. It is proven that corrosion behavior is dependent on multiple factors, such as the properties of crude oil, flow rate, phase inversion, water cut, and pipe material (Ruzic et al., 2006; Zheng et al., 2008). Obviously, the understanding of the corrosion mechanism of production fluids is crucial for the design of oil production and transportation pipelines.
It is challenging to predict flow patterns, the entrainment of water by oil, and water-wetting behavior. Although numerous predicting models and experimental studies have been proposed by researchers worldwide, the corrosion mechanism of multiphase flow and hydrodynamic characteristics of oil-brine mixtures are still not fully understood. This paper aims to provide a comprehensive review of the literature concerning the experimental study of multiphase flow on corrosion and water-wetting modeling and highlight the key parameters that need to be further studied.
2 Internal corrosion mechanism of pipelines caused by water wetting
2.1 Corrosive media
Recently, most studies have primarily focused on the corrosion behavior of the substitute fluids of crude oil on mild steel. The produced water in crude oil is highly corrosive with dissolved corrosive media. The mechanism of some typical corrosive media, such as carbon dioxide (CO2; Li et al., 2006; Yang et al., 2012; Barker et al., 2014), hydrogen sulfide (H2S; Ming et al., 2007; Nešić et al., 2008), and Cl− (Jiang et al., 2012; Zhang et al., 2016), as well as some microorganisms are widely researched. The researches referred in this section are mostly studied in the pipes of carbon steel, stainless steel, or mild steel. Specific material will be named if it is mentioned in the original references.
Among all those corrosive media dissolved in the produced water of crude oil, the most representative ones are roughly summarized as follows.
2.1.1 Corrosive gas
2.1.1.1 H2S
H2S generates from the sulfur-bearing oilfield, especially after the injection of high-temperature steam during the exploration of high-viscosity crude oils (Lin et al., 2014), and is also the most noteworthy metabolite of sulfate-reducing bacteria (SRB). H2S dissolved in water is highly corrosive and deteriorates the corrosion problem (Nesic & Sun, 2010). The ionization reaction is shown below:
Steel gets corroded as the reactions on the steel surface happen (Meresht et al., 2011):
Anodic reaction:
Cathodic reaction:
The released hydrogen ions are strong depolarizing agents and can easily capture electrons at the cathode. In that case, it will result in the pitting corrosion on the metal surface by accelerating the dissolution of the iron at the anode. FeS, among all the productions of the reaction above, such as FeS2, Fe3S4, and Fe9S8, has a relatively high density and agglutinating value. It can form the corrosion production film, which can isolate the steel surface from further corrosion (Nädler & Mewes, 1997).
There is also another type of corrosion induced by H2S. Some hydrogen atoms dissipate in the form of gas, whereas the other atoms are attached to the surface of steel (API 5CT P110 steel) with sufficient energy and permeate into the steel substrate (Simoni et al., 2017). In the progress of hydrogen intrusion, the atoms will be captured and stored in material defects, such as planar defects, crystal dislocations, and stress concentration area with the API 5L-X52 steel (Gonzalez et al., 1997). The pressure increases at the defects as the hydrogen assembles and makes the material [aluminum alloy (Al-Cu-Mg)] mechanical properties such as toughness and ductility decrease, leading to hydrogen embrittlement or the so-called stress corrosion (Alexopoulos et al., 2017) at low tensile stress. Koyama et al. (2012) have studied the influence of hydrogen embrittlement on the properties of alloy (Fe-18Mn-0.6C TWIP steel), as shown in Figure 1.
Influence of hydrogen embrittlement on the properties of alloy: (A) ductility degradation in a hydrogen-charged Fe-18Mn-0.6C TWIP steel and (B) scanning electron micrograph showing an intergranularly fractured surface.
The initial strain rate is 5.1×10−5 s−1. The alloy composition is in wt% (Koyama et al., 2012). Reproduced with permission from Elsevier.
The other influencing factors related to the corrosion rate of H2S, such as temperature, H2S partial pressure, pH, flow rate, Cl− concentration, CO2 partial pressure, and pipeline material, will not be discussed in detail here.
2.1.1.2 CO2
Recently, due to the wide application of the enhanced oil recovery (EOR) in oil production, the fluid becomes more aggressive to the pipeline. As one of the three primary techniques of EOR, gas injection, which generally acidifies the production fluids by the presence of acid gas (Wang et al., 2014b), such as CO2 and H2S, accounts for nearly 60% of EOR production in the United States (Sheng, 2010). It is also a process that introduces miscible gas into the reservoir.
Besides, the CO2 in the produced water of crude oil is commonly generated during the complicated geochemical process. Its corrosion mechanism is basically hydrogen depolarization corrosion whose reaction may destroy the protective film (Xia et al., 1989; Nyborg, 1998).
Currently, there is no common agreement of the mechanism of CO2 corrosion. Davies and Burstein (1980) proposed that the reactions at anode in the CO2 solution are
On the contrary, De Waard and Milliams (1975) presented that the reactions should be
and the reactions above can be simplified as
These reactions are applicable to the exposed metal surface, which would be covered by the corrosion product scales, and the reactions would be as follows after the formation of corrosion product scales:
The mechanism of corrosion product layer formation has been explained (Ingham et al., 2012; Zhang et al., 2012). Based on the Kermani and Morshed (2003) study, the corrosion product scales can be classified as four primary types as shown in Table 1. The effects of the various parameters on scale protectiveness are shown in Table 2 (Kermani & Morshed, 2003).
Characteristics of corrosion product scales.
| Corrosion product scales | Temperature range of formation | Characteristics | Growth habit | Composition |
|---|---|---|---|---|
| Transparent | Room temperature or lower | Thickness <1 μm, transparent; very protective | Forms fast as temperature reduces (room temperature or lower) | Fe and O |
| Iron carbide | – | Thickness <100 μm, metallic, conductive, and not adherent | Spongy and brittle | Fe and C |
| Iron carbonate | 50–70°C (minimum required in laboratory conditions) | Protective, adherent, and not conductive | Cubic morphology | Fe, C, and O |
| Iron carbide+iron carbonate | Maximum 150°C (higher temperature not studied) | Depends on how FeCO3 is blended with Fe3C | – | Ferrous carbide and ferrous carbonate |
Effects of the various parameters on scale protectiveness.
| Parameter | Surface coverage | Stability | Adhesion | Density |
|---|---|---|---|---|
| Oxygen | – | * | * | – |
| Magnesium | – | * | * | – |
| Temperature | * | – | * | – |
| PH | * | – | * | * |
| Carbide phase | – | * | * | – |
-
*, Improvement in certain characteristics of the scale; –, no improvement was observed.
Generally, CO2 and H2S exist in the oil well simultaneously. These two kinds of gas can influence the corrosion rate of each other in both ways. With the same concentration, the solution that contains CO2 and H2S is more corrosive than the solution that dissolves CO2 and H2S, respectively. At 220°C, the corrosion rate of H2S increases as the mass fraction of CO2 goes up and reaches the maximum when the mass fraction of CO2 is 30% (Ramanarayanan & Smith, 1990). There are also studies about corrosion under the mutual influence between CO2 and H2S (Pots et al., 2002; Agrawal et al., 2004; Huang et al., 2015). It is believed that the presence of the hydrocarbon phase was accounted through a so-called water-wetting factor in the study of the CO2 corrosion model (Srdjan et al., 2005).
2.1.1.3 Dissolved oxygen
Generally, the oxygen in crude oil is introduced by the injected water. Oxygen dissolved in water is highly corrosive as H2S. Serious corrosion will occur when the oxygen concentration reaches 0.1 mg/l (Lee et al., 1993). Oxygen corrosion is commonly seen as the corrosion of the oxygen concentration cell under the deposit and the corrosion product scales (Huang & Zhang, 2005). The depolarization corrosion accelerating the corrosion rate happens when the oxygen concentration is low. There is a critical value of the oxygen concentration in exceed of which the corrosion rate drops as the oxygen concentration rises.
2.1.2 Dissolved salts
The dissolved salts in crude oil have a significant impact on the corrosiveness of the produced water. Water conductivity will be greatly enhanced by the dissolved salts, promoting the interaction between anions and cations at a further distance from the pipe surface and accelerating the corrosion rate. Then, the increasing water salinity will reduce the stability of the deposited colloid, such as Fe(OH)2, reducing the quality of the protective film. The amount of the dissolved oxygen will decrease as the salinity exceeds a certain value, weakening the reaction at the cathode and reducing the corrosion rates (Jiang et al., 2012).
The most common dissolved salts in crude oil are chloride, sulfate, and bicarbonate. Generally, the corrosiveness of different anions and cations, among which the chloride and sulfate are supposed be highlighted, differs at low concentration. The chloride ion not only causes stress corrosion but also hampers the formation of oxidation film on the material surface, leading to pitting corrosion. The tendency of the pitting corrosion will increase as the chloride ion concentration rises. The pitting corrosion features the self-catalysis between the large anode and the small cathode. The region around the pitting is protected by the cathode. The corrosion rate can be further deteriorated especially when the pitting area is small, increasing the area ratio between the anode and the cathode (Davies & Burstein, 1980).
Eliyan et al. (2012) assumed that the corrosion behavior exhibited a cathodic dependence on bicarbonate and temperature where the corrosion rates consequently increased. Chloride prevented stable passivation and increased the anodic sensitivity.
2.1.3 Microbiologically influenced corrosion (MIC)
Due to the characteristics of flow media in product oil pipelines, the microbial corrosion behavior is complex. It is unclear how the corrosion process changes in different fluid flow conditions (Song et al., 2016). The corrosion rate of pipeline steel increases attributing to the synergies between the metal surface, abiotic corrosion products, and bacterial cells and their metabolic products (Beech & Sunner, 2004).
It is widely accepted that bacterial culture in artificial growth media is the standard technique to evaluate bacterial population. Harrigan and McCance (1976) introduced the classical methods of using agar-solidified media for the estimation. Particularly, it is necessary to carry out the test for hydrocarbon-oxidizing organisms under some specific situations as described (Bushnell & Haas, 1941).
Corrosion in petroleum product pipelines are enhanced either directly or indirectly by the large numbers of various types of microorganisms (Maruthamuthu et al., 2011). Microorganisms can change the environment conditions (for instance, oxygen concentration and pH value) of the localized metal, which forms the concentration cell on the iron surface and lead to localized corrosion (Eckert, 2015). Microbial corrosion is part of the electrochemical corrosion but differs from other types of electrochemical corrosion. The physicochemical properties of the corroded material (API 5L-X60) will be changed by the microbial metabolism (Song et al., 2016). Because of the nonnegligible influence of the water content and oxygen concentration on the growth of both aerobic and anaerobic bacteria, microbial corrosion is commonly observed where the flow stagnates (Urquidi-Macdonald et al., 2014). A much higher corrosion rate resulting from the synergic mechanism of aerobic and anaerobic bacteria is reported when compared to the one under the respective influence of aerobic and anaerobic bacteria on the pipe of Q235 steel (Chen et al., 2011). Regarded as the main culprits of MIC conventionally in the oilfield system (Guan et al., 2013), several typical bacteria are highlighted by the National Association of Corrosion Engineers (NACE) (2004).
As the main cause of anaerobic corrosion, SRB are the most studied bacteria with the greatest influence on corrosion. In the anaerobic or anoxic environment, SRB uses organic compounds absorbed on metal surfaces as carbon sources producing highly corrosive H2S through sulfate reduction.
Biofilm that consists of extracellular polymeric substance (EPS) and inorganic sediment is formed during the metabolic process of SRB (Lang et al., 2009; Yang et al., 2010). EPS has high viscosity, making the microbial absorption on metal surface an irreversible process. Such bacteria are defined as sessile bacteria (SB) that are attached to surfaces and live in biofilms (Javed et al., 2015). More than 90% of the observed bacteria have the tendency to grow on the surface with EPS produced. Experiments show that the corrosion rates of SB differ on different alloy surfaces (Ping et al., 2013; Guan et al., 2017). Element Ni in the metallic matrix is proven to be positive for the growth, absorption, and metabolic processes of SRB (Lopes et al., 2005, 2006). The mixture of SRB and iron-oxidizing bacteria (IOB) has a synergistic effect, enhancing the pitting corrosion of the coupon (Liu et al., 2015).
2.1.4 Corrosion inhibition
It is accepted that there are two most common sources of corrosion inhibition, namely, corrosion inhibition from crude oil and artificial corrosion inhibitors.
A study showed that inhibitor performances are impaired by increasing precorrosion time and increasing temperature. The resulting corrosion attack is localized within deep pits. The detrimental effect is influenced by both the nature of the steel and the inhibitor composition (Gulbrandsen et al., 1998).
Generally, there is a larger amount of surface active compounds in immature crude oil than in the mature one. Experimental results show that the immature hydrocarbons dispersed in the water phase behave as corrosion inhibitors and can reduce the corrosion rate more evidently than do mature hydrocarbons (Lotz et al., 1991). In some cases where water cuts are extremely high (e.g. 99%) and pipe surfaces are only exposed to water phase, corrosion rates are not high as expected (Efird et al., 2004). A series of studies have reported that these beneficial compounds behave like corrosion inhibitors in corrosion (Stroe et al., 2011). It is confirmed that paraffin in crude oil has an influence on the corrosion of mild steel (Yang et al., 2012). The organic compounds can adsorb onto the pipe surface, forming protective films that isolate the surface from aggressive species in the water phase (Mendez et al., 2001) and even modify the corrosion product layers (Richter et al., 2014).
The beneficial effects of corrosion inhibition are discussed in detail elsewhere (Hernandez et al., 2002, 2006). It is pointed out that the precipitated asphaltenes in crude oil can alter the wettability of the steel surface (Ajmera et al., 2011). Waxes that contain high molecular weight paraffin are supposed to precipitate below its wax appearance temperature. However, localized corrosion on the steel surface will result from the removal of wax film by shear (Yang et al., 2012).
2.2 Influence of multiphase flow
By numerical simulation and experiments, it is confirmed that a small amount of water tends to accumulate at low elevation sections along the long-distance oil transporting pipeline (Xu et al., 2012; Song et al., 2016). However, experimental results may deviate from field flow conditions because both the test fluid and the experimental pipe material are different from the actual oil pipeline.
Currently, the studies of internal corrosion of crude oil pipelines is mostly focused on the corrosive species, whereas the influence of multiphase flow on corrosion does not draw enough attention and remains unclear. Compared to the intuitive gas-liquid two-phase flow, the flow patterns of oil-water two-phase flow are more difficult to identify with less experimental data. Various flow patterns are obtained depending on the fluid properties, pH, pipe material, experimental conditions, and so on. No agreements on the identification of flow patterns, formation of slug flow, and viscosity prediction of different phases have been reached.
2.2.1 Influence of multiphase flow on corrosion
Water phase with dissolved corrosive media mentioned above is directly related to internal corrosion. Droplets with small diameters aggregate into larger droplets at a certain velocity and water cut, which are influenced by the oil-water interfacial tension and the different densities of the phases in fluid. Droplets dispersed in oil start to move downward when the turbulent energy of flow is not sufficient enough to suspend droplets forming stream at the bottom of pipe. Then, a scenario known as “water wetting” happens when water droplets come into contact with the pipe surface, increasing the likelihood of internal corrosion (Cai et al., 2004).
The internal corrosion is influenced by numerous factors of multiphase flow and three noteworthy aspects are roughly summarized as follows (Pots et al., 2006).
2.2.1.1 Distribution of phases in fluid
What matters most is the distribution of water phase and the wettability of pipe wall. A variety of flow patterns or flow regimes may originate from the way how the fluid phases physically distribute themselves within the tube (Shoham, 2006). The observed flow patterns are the results of a number of parameters (Taitel & Dukler, 1976), such as phase content, fluid properties, and pipe geometry. Microbial corrosion becomes a concern when solid impurities occupy the bottom of the pipe, aggravating the corrosion caused by dissolved oxygen. Corrosion inhibitors dispersed in fluid may not be capable of reaching the place where corrosion is expected to be controlled if they are not chosen properly. The details on the mechanism of phase wetting will be discussed in the later sections.
2.2.1.2 Mass transport of species
The species include the corrosion product, corrosive medium, solid particles, corrosion inhibitor, and so on. The distribution and properties of corrosion inhibitors along the pipeline are always crucial issues to the pipeline operators (Gece, 2008; Raja & Sethuraman, 2008). For certain corrosion prediction models of protons, organic acids, and oxygen, the key step is the calculation of the mass transport of corrosive species across the diffusion boundary layer to the wall (Nešić, 2007). In the case that a corrosion product scale forms, protective corrosion product scales will be hampered to form by high mass transfer. Then, the key step of corrosion prediction models turns to the removal of the corrosion product from the pipe wall.
2.2.1.3 Flow-induced corrosion
The protective corrosion product films (iron carbonate or iron sulfide scales), pH stabilization, and corrosion inhibitors may get destabilized and flushed away from the pipe wall by shear stress induced by high fluid velocity (Lévesque et al., 2008; Wang et al., 2016). The impact force of liquid droplets and solid particles are also the main reasons of flow-induced corrosion (Wang et al., 2014a). Thaker and Banerjee (2016) presented the distinct regimes and transition boundaries of erosion-corrosion and summarized the inducements as four distinct erosion-corrosion phenomena: flow-accelerated corrosion, shear stress-induced erosion, liquid impact-induced erosion, and cavitation erosion. The details are presented as follows.
Flow-induced corrosion or erosion corrosion is known as the phenomenon of metal damage led by the high velocity of fluid. It is an outcome of the synergy of electrochemical corrosion and mechanical erosion and derives from the coupling effect of fluid flow and corrosion. Considering that flow-induced corrosion is closely related to the flow patterns of fluid, it does not simply emerge from the superposition of flow and corrosion but the self-catalytic effect (Wan et al., 2017), namely, corrosion accelerates erosion and erosion reversely works on corrosion. The situation is especially salient, where fluid velocity abruptly changes as the flow geometry varies, such as obstruction or protrusion in the pipe (Nesic & Postlethwaite, 1990). Zhao et al. (2016) deduced the regularity of local erosion-corrosion rates under different impact angles and assumed that corrosion behavior is deeply affected by resistance to the deformation of steel surface. Erosion-corrosion also significantly attributes to the breakdown of pumps, valves, and pipes in the industries of petroleum, water conservancy, hydropower, and chemistry. It is one of the tremendous risks that feature locality and abruptness, attributing to the unplanned shutdown of petroleum and petrochemical corporations (Jin et al., 2017).
When fluid flows slowly, dew point corrosion and under-deposit corrosion may happen where settling deposits narrow the local pipe diameter, increasing the velocity of relevant position. On the contrary, protective scales on the internal pipe wall could be destroyed by shear stress induced by the high velocity of fluid. There exists a range of velocity in which the pipe stays safe comparatively and a breakaway velocity exceeding which the material loss increases dramatically especially for the material with films on the surface.
Solid particles in crude oil accompany flow-induced corrosion (solid particle erosion) in liquid-solid two-phase flow. Solid particle erosion is greatly influenced by particle properties such as shape, size, density, and hardness.
Particle size is reported to play an important role in the erosion magnitude as well as particle shape according to Levy and Chik (1983). Gandhi and Borse (2002) studied the cast iron erosion behavior under the impact of different particle sizes with two different impact angles.
Compared to smaller particles, larger particles possess larger kinetic energy at the same velocity when they lash the material surface. In general, the higher the density and hardness of particles with uniform size and shape are, the higher the erosion rates become. This may vary depending on the hardness of the material (Antonov et al., 2016). Higher density increases the kinetic energy, the impact force of particles, and the erosion rate. Considering the influence of flow patterns on particle distribution, it should be noted that the local particle concentration could be unexpectedly high even when the overall particle concentration is low, increasing the erosion magnitude. However, the shielding effect observed by Chen (2004) can reduce the erosion rates when interactions among particles get stronger as the local particle concentration exceeds a certain value. Particle size has no obvious changes through the whole experiment, but particle roughness is the reason why erosion-corrosion rate drops as experiment goes on. Corrosion associated with flow is summarized in Figure 2.
Classification of corrosion associated with flow.
2.3 Phase wetting behavior
2.3.1 Properties of crude oil
2.3.1.1 Classification of crude oil
Crude oil mainly consists of a variety of hydrocarbons, such as paraffins, aromatic compounds, and naphthenic compounds, stored in reservoir. There are two typical methods broadly accepted to classify crude oils: the American Petroleum Institute (API) gravity and the proportions of the organic compounds. The API gravity is a parameter that changes in inverse proportion to the density of crude oil.
Efird et al. (2004) confirmed that, according to a detailed identification of the roles of chemical compounds in corrosion, the compounds may present a profound effect on the wettability of crude oil and the chemistry of the brine solutions.
Richter et al. (2014) classified the categorization of crude oils based on their ability to inhibit corrosion and alter the steel wettability, as shown in Table 3.
Categorization of crude oils.
| Categorization | Corrosion inhibition | Alteration of the steel wettability |
|---|---|---|
| Category I | Positive | Positive |
| Category II | Positive | Negative |
| Category III | Negative | Positive |
| Category IV | Negative | Negative |
It is verified that some organic components (mostly asphaltenes) in crude oil can form highly viscoelastic films at the interface between oil and water (Horváth-Szabó et al., 2005; Pawar et al., 2011), slowing down the coalescence of oil drops (Gaweł et al., 2015).
2.3.1.2 Density
As proven by Xu (2007), the entrainment of water in crude oil is dependent on the density and viscosity of crude oil.
Heavier oils are generally more protective than lighter ones with regard to corrosion (Lotz et al., 1991). Higher molecular weight oil reduces the corrosiveness of water-in-crude oil emulsions (Craig, 1998), but the influence of the oil on corrosion rates has not been quantified. The great difference between the oil and water densities makes it easy for the heavier water to accumulate at the pipe bottom. On the contrary, oil and water of similar densities in a two-phase flow system cannot be separated because of natural coalescence (Kee, 2014). The term “critical water break value” Wbreak is defined by De Waard et al. as the water content that can be entrained by oil phase (Smith et al., 2003). Then, an empirical linear formula based on field data analysis is established to estimate the entrainment of water in crude oil:
The water phase is fully entrained by oil when water cut in an oil-water system is below Wbreak, avoiding the water wetting of the steel surface.
2.3.1.3 Viscosity
The viscosity of crude oil is another key property that plays a crucial role in petroleum engineering such as the evaluation of fluid flow, simulation of reservoir, and design of production facilities. Water dispersed in oil is easier to be entrained by more viscous oil. Kokal (2002) explained comprehensively the non-Newtonian behavior in oil-water systems. The apparent viscosity of the mixture is pretty higher than the viscosity of the respective phase when the fluid is highly emulsified. The developed prediction models and determination of viscosity have been developed by numerous researchers (El-hoshoudy et al., 2013; Hemmati-Sarapardeh et al., 2014; Sandor et al., 2016; Hajirezaie et al., 2017).
2.3.1.4 Interfacial tension and surface tension
The wettability of steel by water or oil can be evaluated by the interfacial tension γwater-steel and γoil-steel. The γliquid-steel values of different liquids certainly differ from each other; for example, the γoil-steel can be different depending on the types of crude oils. However, for the mixture of water and oil, γwater-steel is independent of the properties of crude oils (Smith et al., 2003). The lower value of interfacial tension between one phase and steel indicates the better wetting of steel by the same phase. The following equation, where γwater-steel<γoil-steel (γwater-steel is a constant for the same steel), explains why steel is wetted better by water than oil when the two phases are well emulsified in a dispersed flow considering the characteristics of water wetting on steel (Smart, 2001).
The interfacial tension is also an important factor to evaluate the stability of the emulsion because the γwater-oil value is the energy released when water-in-oil demulsification happens. The higher value of γwater-oil means more energy stored in the emulsion, which is easier to break; that is, the lower value of γwater-oil ensures tighter emulsion.
Furthermore, this can explain why heavier oil is more protective than lighter oil to the pipe (Mendez et al., 2001), as γwater-oil decreases as the density of crude oil increases, making the water-in-oil emulsion more stable and oil more likely to wet steel surface than water.
The effect of surface tension, flow rate, and viscosity on the droplet size is observed through both experiment and calculation by Chen et al. (2016).
2.3.1.5 Contact angle
The schematic diagram of contact angle θ of phase 1 in two-phase flow is shown in Figure 3. According to the Young (1805) equation, based on the balance of the horizontal components of the interfacial tensions at equilibrium, the contact angle θ in oil-water two-phase flow can be measured by the interfacial tension:
Measurement of contact angle θ.
where γso, γsw, and γwo are the interfacial tensions of steel-oil, steel-water, and water-oil, respectively.
The wetting behavior of the liquid droplet is controlled by the balance of the cohesion force inside the liquid drop and the adhesion force between the solid surface and liquid droplet. When the cohesive force is stronger than the adhesive force, the water droplet is less likely to spread on the nonwettable solid surface (hydrophobic surface) and forms at equilibrium a spherical cap with a contact angle of larger than 90° (Figure 4A). In contrast, when the solid has a high affinity for water, in which case the cohesive force is not stronger than the adhesive force, water spreads on the hydrophilic surface (which is preferentially wetted by water) with a contact angle of less than 90° (Figure 4B).
Wettability of surface: (A) water droplet beads up on a hydrophobic surface and (B) water droplet spreads on a hydrophilic surface.
Gaweł et al. (2016) found the relationship between the value of the contact angle and the formation of particle stabilized emulsions. It was reported that the contact angle can be changed by prewetting the steel surface with surface active compounds in crude oil or corrosion inhibitors, altering the surface wettability (Tang, 2011).
It should be noted that contacting and wetting are two different processes. The effects of the contact angle on corrosion have been investigated by a number of researchers (Ajmera et al., 2011; Ayello et al., 2011). The classification of wettability is summarized in Table 4.
Classification of wettability.
| Contact angle | Degree of wetting | Strength of |
|
|---|---|---|---|
| Solid-liquid interactions | Liquid-liquid interactions | ||
| θ=0° | Perfect wetting | Strong | Weak |
| 0°<θ<90° | High wettability | Strong | Strong |
| Weak | Weak | ||
| 90°≤θ<180° | Low wettability | Weak | Strong |
| θ=180° | Perfect nonwetting | Weak | Strong |
2.4 Phase inversion point (IP)
2.4.1 Definition of water phase inversion
For crude oil containing water, water is entrained by continuous oil in the form of dispersion at low water cut. As the amount of dispersed water increases, the area of interface between water and oil grows, increasing the cohesive force between water drops. Hence, the viscosity is raised by interfacial energy. When the water cut rises to a critical value, a scenario named phase inversion happens, where the water dispersed in oil (w/o) emulsion turns to oil dispersed in water (o/w) emulsion. This transition is also associated with abrupt changes of the rates of momentum, heat, and mass transfer between the continuous and dispersed phases and between the dispersion and the system solid boundaries (Shoham et al., 1989). The critical water cut is called phase IP (Angeli & Hewitt, 2000a; Xu, 2007; Kumara et al., 2009; Yusuf et al., 2012) or defined as emulsion IP (EIP; Fingas & Fieldhouse, 2004; Papavinasam et al., 2007).
The viscosity of the oil containing water is mainly determined by the viscosity of crude oil, the temperature of the fluid, and the water cut. Non-Newton rheological behavior presents when the water cut is high but the fluid still follows the constitutive equations of power-law fluid. As a function of water cut, the apparent viscosity increases with water cut before IP, increasing the energy needed to transport the fluid. The apparent viscosity reaches peak value at the IP and decreases as water cut grows continuously.
This is a complex phenomenon that depends on several factors such as physicochemical fluid properties among others (Arirachakaran et al., 1989; Brauner & Ullmann, 2002). If crude oil is transported by continuous water phase, resistance to flow can be decreased, increasing the transportation efficiency. However, if a pipeline operates with water cut values similar to or larger than IP, the continuous water phase will lead to water wetting of the pipe surface. Operators of pipelines are suggested to avoid operating the pipe in the flow conditions around IP.
2.4.2 Phase inversion modeling
Selker and Sleicher (1965) determined the variation range of IP of different experimental conditions and assumed that the liquid viscosity ratio is the primary factor that affects the limitation of the range with the wetting properties of the container material, liquid densities, and surface tension being the subfactors. The tendency of oil to be dispersed increases as the oil phase viscosity grows, increasing both the minimal oil volume fraction that can be continuous and its maximal volume fraction that can be dispersed.
Arirachakaran et al. (1989) proposed an empirical correlation to calculate the IP:
where ηo is the viscosity of oil phase (Pa·s).
As shown by the correlation, the water cut required to invert a dispersion decreases as the oil viscosity increases. They also present two pressure gradient models that match well with the experimental data and find that pressure gradients increase significantly at the phase IP as substantiated by Angeli and Hewitt (1996).
Assuming that there exists a negligible interfacial shear and no-slip between the two layers, Nädler and Mewes (1997) presented an correlation based on the momentum equations for stratified flow:
where k1, k2 are empirical parameters, Co, Cw and no, nw are the parameters of the Blasius equation for the friction factor, D is the pipe diameter, ρo, ρw are the densities, and ηo, ηw are viscosities of the pure oil and water phases, respectively. It is suggested that k1 reflects the wall-liquid contact perimeter, as determined by the in situ configuration, and k2 accounts for the flow regime in each of the phases.
According to the studies of a number of researchers (Xu et al., 2010; Zhang et al., 2015, Zhang et al., 2017), no agreement on the mechanism understanding of phase inversion has been reached, and for the phase inversion of the more complicated oil-gas-water three-phase flow, a few scholars have initiated the preliminary study (Kokal, 2002; Perazzo et al., 2015).
3 Studies on water-wetting modeling
Wetting refers to the study of how a liquid deposited on a solid (or liquid) substrate spreads out. The earlier studies do reveal the characteristic of water entrainment mechanism. Meanwhile, it has to be admitted that these criteria are somewhat crude, neglecting the impact of varying properties of fluid, the flow regime, and the flow geometry. Water wetting is closely related to the origination of corrosion on steel surfaces in an oil-brine mixed condition (Cai et al., 2012; Larsen, 2013). The processes of water separation from the emulsion have been explained by Abdel-Raouf (2012).
The current models of describing the likelihood of water wetting have their own limitations, as they cannot involve all the related factors considering the synergetic effects of multiple parameters. The complicacy of multiphase flow is neglected when only one aspect, such as the velocity of fluid, shear stress, transitions of flow patterns, is taken into consideration to model corrosion risks. However, a much more complicated model involving too many parameters does not mean higher accuracy and practicability, especially when a parameter of the mechanism is still controversial.
Water tends to occupy the pipe bottom or lower area induced by the changes of terrain elevations, giving rise to water wetting. The probability of internal corrosion can be significantly reduced if water dispersed in continuous oil can be fully entrained avoiding contact with the pipe wall (Lotz et al., 1991; Pouraria et al., 2016). The water droplet is suspended and fully dispersed in oil only when the turbulent velocity of oil is high enough, breaking continuous water phase into small droplets and counteracting the gravity force. Otherwise, the likelihood of water wetting is depending on the wettability (hydrophilic or hydrophobic) of the pipe surface, in competition to wetting by oil, which prevents corrosion (Paolinelli & Nesic, 2016). A free layer of water forms on the pipe surface with the turbulent velocity of fluid being slow and the pipe surface staying hydrophilic.
3.1 Parameters related to water-wetting modeling
As discussed in Section 2, the effects of various corrosive media on the corrosion of steel pipes have been studied extensively by establishing corrosion models, but the prediction of water wetting on these corrosion rates has always been an elusive problem. The parameters related to water-wetting modeling are discussed as follows.
3.1.1 Maximum diameter dmax and critical diameter dcrit
As discussed in the former chapters, the internal corrosion of pipes mainly depends on the phase that wets the pipe. The droplet size of the dispersed phase is up to its type of dispersion (Brauner & Ullmann, 2002).
Under the experimental conditions, the liquid drops are broken up into smaller droplets by the turbulent energy and shear stress introduced into the system by the pump. However, according to the properties of the fluid, the drop breakage rate and the resulting drop size distribution may differ a lot from the coalescence rate of different dispersed phases when the water cut is below and above IP, respectively. There exists a dynamic equilibrium, which must be kept for a stable liquid dispersion, between the two competing rates of breakage and coalescence. Droplet size distribution is determined by the drop breakage rate, and it is especially crucial for the estimation of water wetting (Tsahalis, 1977; Brauner, 2001; Xu et al., 2011).
For the oil-water system with a water cut lower than 5%, the maximum diameter of stable water droplet without further breakup can be obtained by the correlation (Hinze, 1955):
where ε is the mean energy dissipation rate per unit of mass of the oil phase (W/kg) and σ is the interfacial oil-water tension (N/m).
However, the calculated values of dmax are lower than the experimental values, as the disruptive energy of oil phase decreases as the fraction of the dispersed water phase increases. For the situation of the water cut of more than 5%, the model presented by Brauner (2001) is applicable:
where CH is a constant that is approximately equal to 1 and εw is the water cut. The calculated value dmax agrees well with flow loop experimental data when H≈1.8 (Paolinelli & Nesic, 2016). It should be noted that both models above are effective for a limited length of experimental flow loop.
The cumulative volume fraction V of droplets with diameters larger than d can be calculated by a Rosin-Rammler-type equation after dmax is obtained (Karabelas, 1978):
The critical diameter dcrit of the water droplet is the critical size of diameter that can be suspended by the force induced by turbulence. When dmax is larger than dcrit, the suspension force induced by turbulence is not strong enough to counteract the gravity of the water droplet, and water coalescence happens, forming a free water layer on the pipe surface. Hence, the critical diameter dcrit is an important criterion to judge the wetting conditions and estimate the corrosion risks.
Barnea (1987) assumed that the critical droplet size dcrit is determined by the following correlation:
In the horizontal oil-water flow, the critical droplet size dgravity is dominated by the suspension and gravity force, which can be calculated by the following equation:
d deformation is the critical droplet size determined by the deformation force of nonspherical droplets, which can be expressed as
where β is the inclination angle of the pipe,
The critical diameter dcrit can be obtained by solving the equations of suspension force and gravity force (Paolinelli & Nesic, 2016):
where ρw and ρo are the densities of water and oil, respectively; CD is the drag coefficient of the water droplet; and ν′ is the vertical turbulent velocity fluctuations in the continuous-phase flow (m/s).
3.1.2 Critical velocity Vcrit
Critical velocity Vcrit is obtained when the water droplet is exactly suspended by turbulence energy, namely, dmax equals dcrit, and can be applied as a criterion to judge the stability of water-in-oil dispersion. If the fluid velocity is higher than Vcrit, it can be justified that water is fully dispersed in the oil and no water wetting exists. On the contrary, if the fluid velocity is lower than Vcrit, there is a higher tendency of moving downward of water droplets with a diameter larger than dcrit, resulting in water wetting (Brauner, 2001; Cai et al., 2004).
Russell et al. (1959) found that some oils with velocities larger than 1 m/s can sweep out settling water when the water cut is no more than 20%. In the upward inclined pipe flow, there may exist the presence of a countercurrent water flow; at the bottom of the pipe, the velocity of oil-water fluid is low (Vigneaux et al., 1988; Flores, 1997).
3.1.3 Critical inclination angle
Greatly influenced by the fluid velocity, the critical inclination angle is another key parameter of the prediction models of water entrainment. The likelihood of water coalescence is low and no water layer forms when the pipe inclination is smaller than the critical inclination angle.
3.2 Water wetting models
It is crucial to predict the initiation of oil/water wetting in the corrosion evaluation. In the case of water wetting, the corrosion rate can be estimated by the key factors of free water layer such as its velocity and shear stress with wall, which influence the mass transfer between the pipe and the bulk flow (Nešić, 2007).
Among all those key factors, the determination of critical velocity has always been a hot issue for the estimation of water entrainment. A number of researchers have established prediction models of critical velocity in different flow patterns (Srdjan et al., 2005; Pouraria et al., 2016). However, due to the complexities of the hydrodynamic behaviors of multiphase flow, more quantitative researches need to be further developed.
Hinze (1955) assumed that the force balance between the surface tension force and the dynamic pressure force of the turbulent eddies surrounding the droplet can be applied to calculate the maximal droplet size dmax:
where ρc is the density of the continuous phase (g/cm3), σ is the interfacial tension (dyne/cm), and ε is the energy input per unit mass and time or dissipation per unit mass (cm2/s3).
Based on the energy balance between the turbulent kinetic energy from the continuous phase and the surface energy formed in dense dispersion, Brauner (2001) later extended the model of Hinze (1955) for dense dispersion system:
where εd is the input water cut and e̅ is the mean energy dissipation rate.
Wicks and Fraser (1975) proposed a simplified prediction model of critical velocity of oil phase based on the assumption that the correlation for the sand particles could be equally applied to liquid droplets of the same size and the data of the droplets size are from Hinze (1955). However, this model is not applicable for oil with high water cut, as it neglects the process of water coalescence, which is significantly affected by the density and viscosity of the fluid, the interfacial tensions between oil and water, and the pipe diameter. The model of Wicks and Fraser is improved by Wu in 1995 (Yiing-Mei, 1996).
Xu et al. (2011) suggested a model based on the instability of the interface and found that the oil phase is not necessarily turbulent with its Reynolds number (Re) much lower than 2100 when it reaches the critical velocity (for one-phase flow, when Re<2100, the flow is laminar; when 2100<Re<4000, the flow stays in the transition field; when Re>4000, the flow is turbulent). Through experiments, it is confirmed that the backflow of water may occur when the oil velocity is lower than the critical velocity at low sections of the pipe. The critical velocity increases as the pipe diameter grows. An empirical model (Smith et al., 2003) has been proposed based on the previous work of De Waard and Lotz (1993) using an analysis of the emulsion breakpoint. The model takes the API gravity, emulsion stability, and water wetting into consideration but ignores the effect of pipe diameter, oil density, oil viscosity, and system temperature on the critical velocity.
Based on the works of Karabelas (1977), Segev (1985) established conservative models of predicting the water droplet distribution in the pipe in polar coordinate (r, θ):
where co is a constant, α is the ratio between droplet settling velocity and turbulent fluctuation velocity shown as Equation (30), and R is the radius of the pipe.
where Vs is the terminal velocity and D is the droplet eddy diffusivity.
However, the critical velocity attained from the study is lower by having the water droplets displaced by solid particles of the same size in the correlation where the Stokes assumption for the droplet settling may not be applicable.
Tsahalis (1977) presented a model of calculating the critical velocity based on the instability of the water layer and water entrainment by oil.
Cai et al. (2004, 2012) formulated a mechanistic model to predict the critical entrainment velocity required to disperse the water phase based on the liquid-liquid flow pattern transition work by Brauner (2001).
Based on the model of Cai et al., Tang (2011) developed a water wetting that included the effect of surface wettability (change in contact angle). He assumed that additional interactions between the water-steel and oil-steel surfaces in the presence of surface active compounds from crude oils can influence the droplet breakup process and the water entrainment by the flowing oil. The new model accounted for the extra-turbulent kinetic energy required to create new surfaces resulting from the solid-fluid surface interactions. He calibrated and compared the model to laboratory data using different types of crude oils.
Paolinelli and Nesic (2016) proposed criteria to assess corrosion risk in oil-water two-phase pipe flow focusing on the mechanisms of water entrainment and the distribution of water droplet size. This model agrees well with the experimental results by considering the contact angle and phase inversion. However, no further research of oil with a higher water cut of more than 20% has been done because of the lack of corresponding experimental data.
4 Experimental study of multiphase flow
According to field experience, corrosion has been observed in a wide range of water cuts (Craig, 1998). A small amount of water may induce corrosion problems. The phenomenon of the sharp increase in corrosion rate at a certain water cut has been experimentally verified by many laboratory tests (Craig, 1998; Al-Hashem et al., 2000; Papavinasam et al., 2007). This critical water cut of corrosion rate break (CRB) is defined as the critical water break value or the EIP.
In other cases, pipelines that transport production fluids with relatively high water cuts (for example, up to 50 wt%) have been reported to be free from corrosion (Li et al., 2006).
Different from the physical definition of phases in physical forms, the multiphase in the petroleum industry is a mixture of more than two phases of different chemical components such as gas, oil, and water. The current flow patterns observed are influenced by a number of factors as discussed in the previous chapters and the classifications of flow patterns are quite subjective results based on various experimental conditions. Even the same flow pattern may be named differently by different researchers.
The current experiments of the characteristics of oil-water flow vary greatly worldwide, but the most common experiments can be classified as small-scale laboratory tests and large-scale flow loop experiments. The three mostly applied techniques of identifying flow patterns are as follows.
4.1 Observation
The images captured by high-speed video camera and visual inspections by researchers are the most direct and convenient method to record the fluid flow through transparent pipe sections, especially for high-speed fluid flow. However, the apparent flow regimes may not reveal all the hydrodynamic characteristics of multiphase flow.
Wegmann et al. (2007) applied laser-induced fluorescence (LIF) to distinguish the different phases and capture flow patterns images with the help of a high-speed camera. It is found that the decreasing pipe diameter changes the flow pattern maps and the behavior of the transition boundaries. In a pipe diameter of 5.6 mm, the intermittent-intermittent flow is transformed into dispersed-intermittent flow by the increasing speed of oil phase, and in a pipe diameter of 7 mm, the transformation works reversely, which is not recorded in the previous literatures.
Similar to the classification method of flow patterns from the Wegmann et al. work, Bannwart et al. (2009) classified the gas-water and oil-water flow patterns separately (first gas-water and second oil-water) to four different groups: S=stratified, A=annular (core), I=intermittent, and B=bubbles. All observed flow patterns of dense oil-water-gas are indicated on the flow maps of oil and gas superficial velocities for both horizontal and vertical flows.
4.2 Conductivity probes
As the conductivity is an important physical property that can be used to distinguish different phases and record the phase distribution in the space as time goes on, it is convenient to measure the wettability of the pipe surface and help to identify flow patterns by examining the conductivity of a group of probes that are positioned as planned. The merit of this wettability spreading method is that it can be conducted in higher pressure and flow conditions (Wang & Zhang, 2016).
4.3 Electrical tomography
Electrical tomography is based on the measurement of the electrical property of the matter, such as resistivity, capacitance, and impedance. The technique is used to visualize multicomponent fluids moving in a process pipeline. The electrically based tomography system typically comprises three functional systems: sensor system, data acquisition system (DAS), and image reconstruction system (Dong et al., 2003). There are two core sensor systems of electrically based tomography: electrical resistance tomography (ERT) for the resistivity measurement and electrical capacitance tomography (ECT) for the capacitance measurement. Sharifi and Young (2013) have introduced the applications of ERT in detail.
Açikgöz et al. (1992) classified 10 different flow patterns for oil-gas-water three-phase flow by the type of dominating phase, namely, oil based and water based. The classification represents the flow characteristics of oil-gas-water three-phase flow but not recommended to be popularized because of the intricate transitions between oil- and water-based flows.
Angeli and Hewitt (2000b) proposed that the liquid droplet size distribution can be significantly influenced by the pipe material. The droplet size in steel conduit is comparatively smaller than the value in plastic conduit under the same flow conditions. Both the maximum diameters and the sauter mean diameters (SMD) rely on the velocity of the continuous phase. The Hinze (1955) equation is proven to underestimate the maximum drop size of the available experimental results.
With high-speed camera capturing the flow patterns and with conductivity probes recording the phase distribution, Vielma et al. (2008) examined three probabilistic distributions of fully dispersed flows and developed an empirical correlation to predict the SMD of droplets across the pipe cross-section for dispersed oil in water. A number of theoretical correlations for the maximum drop size, such as Hinze (1955), Kubie and Gardner (1977), Angeli and Hewitt (2000b), and Kouba (2003), are proven not to predict the experimental data as accurately as expected.
With the experimental data from Trallero (1995) and models for predicting flow pattern transition from Brauner (2001) for the horizontal system of EoD>>1 (EoD=ΔρgD2/8σ), an experimental oil-water flow pattern map was reported by Ullmann et al., as shown in Figure 5 (Brauner & Ullmann, 2002).
Experimental oil-water flow pattern map based on the superficial velocity of water and oil: 1=neutral stability boundary for smooth stratified flow; E=entrainment of oil drops into the water layer; EU=equal velocity of fluids in stratified layers; LTo=laminar/turbulent transition in the oil layer; 4=H-model, water continuous; 5=H-model, oil continuous; LTm=laminar/turbulent transition, oil continuous phase (Brauner & Ullmann, 2002).
Reproduced with permission from Elsevier.
The general distribution and location of water in a three-phase flow system is summarized in Table 5 (Kee et al., 2015). The corresponding flow patterns are shown in Figure 6 (Kee, 2014).
Water distribution in the horizontal gas-oil-water flow (Kee et al., 2015).
| Flow patterns | Water distribution | Water location |
|---|---|---|
| Stratified (ST) | Separated | Bottom |
| Elongated bubble (EB) | Separated | Bottom |
| Slug (SL) | Dispersed/separated | Mostly bottom |
| Wavy annular (WA) | Dispersed | Mostly circumferential |
| Annular-mist (AM) | Dispersed | Circumferential |
Schematics of oil-continuous horizontal three-phase flow patterns showing stratified, elongated bubble, slug, wavy annular, and annular-mist flows (Kee, 2014).
With the experimental data, a series of flow pattern maps at 1%, 5%, 10%, and 20% water cuts are plotted by Kee, and the flow pattern map at 1% water cut is shown in Figure 7 (Kee, 2014). The different spots representing different flow patterns have been obtained from experiments and the transition boundaries are from water-wetting prediction models. It should be noted that the flow transition boundaries (purple lines) do not represent any absolute value, as the flow pattern transition is not an instantaneous progress. Some experimental data have been found not to strictly obey the transition boundaries but still close to the transition lines. This can be explained by the uncertainties of the predicted transition boundaries.
Comparison of horizontal three-phase flow pattern results with WW model at 1% water cut for the CO2-LVT200 (model oil)-water system (Kee, 2014).
According to the current experimental studies of multiphase flow, the experimental results are not substantially representative of actual engineering conditions. The reasons can be summarized as follows:
Most experiments focus on the horizontal and vertical flows, whereas the pipe elevations under actual conditions are quite complicated.
The droplet size distribution in flow loop experiments is inevitably influenced by the shear rates induced by pumps and not representative of the long-distance pipeline.
Generally, most research institutions use some substitute, such as kerosene and white oil, to replace crude oil, as the large quantity of crude oil is difficult to transport. The substitute oil still cannot completely replace crude oil even if the properties are quite similar.
Some attachment facilities, such as the transparent pipe section for inspection and the conductivity probes fastened to the internal pipe surface, can also influence the wetting behavior of the fluid. The multiphase flow may not be fully developed.
5 Conclusions
This paper aims to discuss the effects of different parameters on the internal corrosion of pipelines in the petroleum industry and provide a comprehensive understanding of the hot issues in the current research. Different empirical and mechanistic corrosion prediction models by different researchers have been presented. According to the overview of the previous researches, the conclusion can be drawn that the internal corrosion of pipelines for crude oil containing water is influenced by various factors. The current prediction models and experimental studies involving interdisciplinary knowledge still need further improvements. The conclusions can be drawn as below.
The produced water in crude oil dissolves multiple corrosive media of which the respective mechanism still holds some aspects unknown. The corresponding corrosion products may be beneficial or harmful depending on the corrosive medium and the properties of material may also be changed as discussed in Section 2. Hence, it is hardly possible to draw a common conclusion of how internal corrosion initiates under the function of different corrosive media.
Some chemical drugs injected into the pipe work as corrosion inhibitors, making the corrosion behavior of steel more complicated to predict. Some corrosion inhibitors can alter the wettability of steel and assure more water to be entrained by oil, preventing water wetting.
Multiphase flow affects internal corrosion through three aspects: (1) distribution of phases in fluid, (2) mass transport of species, and (3) flow-induced corrosion. The third aspect is noteworthy.
The hydrodynamic characteristic of multiphase flow is correlated to the fluid properties (such as the density, viscosity, and crude chemistry), the interfacial behavior, the reactions at steel surfaces, and so on. The mechanistic and empirical models are still worth further investigation into the effects of phase inversion, formation of emulsions, and the modification of corrosion product layers on corrosion risk.
Water-wetting phenomenon and wettability of steel have been explained in this paper. The properties of the fluid and pipe material and the operating conditions are all accountable for the accuracy of water-wetting modeling. The current water-wetting model is somewhat conservative and more precise prediction models are crucial for optimizing production and reducing risk in the future work.
The current flow loop experiment may not conform with petroleum field operating conditions. The oil-water two-phase flow is mostly studied in horizontal and vertical flows at low water cuts and the oil-gas-water three-phase flow still has not reach an agreement. More experimental data, especially at high water cut, are needed for corrosion prediction models in different multiphase flow regimes.
About the authors
Hao Zhang is a PhD candidate in vehicle operation engineering at Beijing Jiaotong University (BJU). He graduated from BJU in 2013 with a bachelor’s degree in mechanical design and automation. His research area focuses on the internal corrosion of pipelines in the petroleum industry.
Hui-qing Lan is a professor and a PhD supervisor at BJU. She completed her postdoctoral work at the University of Tokyo. She holds a PhD in engineering (2002) from the China University of Petroleum (Beijing) and is a member of the Petroleum Storage and Transportation Committee of the China Petroleum Society.
Acknowledgments
This project was supported by the National Key R&D Program of China (no. 2017YFC0805005) and the Special Fund for Quality Inspection Scientific Research in the Public Interest of China (no. 201410027).
References
Abdel-Raouf M. Factors affecting the stability of crude oil emulsions. Crude oil emulsions – composition, stability and characterization. Croatia: Intech, 2012: 183–204.10.5772/2677Search in Google Scholar
Açikgöz M, Franca F, Lahey R. An experimental study of three-phase flow regimes. Int J Multiphase Flow 1992; 18: 327–336.10.1016/0301-9322(92)90020-HSearch in Google Scholar
Agrawal AK, Durr C, Koch GH. Sulfide films and corrosion rates of AISI 1018 carbon steel in saline solutions in the presence of H2S and CO2 at temperatures up to 175F. In: Corrosion 2004, NACE International.10.5006/C2004-04383Search in Google Scholar
Ajmera P, Robbins W, Richter S, Nešić S. Role of asphaltenes in inhibiting corrosion and altering the wettability of the steel surface. Corrosion 2011; 67: 105006–105011.10.5006/1.3651015Search in Google Scholar
Al-Hashem AH, Carew JA, Al-Sayegh A. The effect of water-cut on the corrosion behavior of L80 carbon steel under downhole conditions. In: Corrosion 2000, NACE International.Search in Google Scholar
Alexopoulos ND, Charalampidou C, Skarvelis P, Kourkoulis SK. Synergy of corrosion-induced micro-cracking and hydrogen embrittlement on the structural integrity of aluminium alloy (Al-Cu-Mg) 2024. Corros Sci 2017; 121: 32–42.10.1016/j.corsci.2017.03.001Search in Google Scholar
Angeli P, Hewitt GF. Pressure gradient phenomena during horizontal oil-water flow. New York, NY: American Society of Mechanical Engineers, 1996.Search in Google Scholar
Angeli P, Hewitt G. Flow structure in horizontal oil-water flow. Int J Multiphase Flow 2000a; 26: 1117–1140.10.1016/S0301-9322(99)00081-6Search in Google Scholar
Angeli P, Hewitt GF. Drop size distributions in horizontal oil-water dispersed flows. Chem Eng Sci 2000b; 55: 3133–3143.10.1016/S0009-2509(99)00585-0Search in Google Scholar
Antonov M, Pirso J, Vallikivi A, Goljandin D, Hussainova I. The effect of fine erodent retained on the surface during erosion of metals, ceramics, plastic, rubber and hardmetal. Wear 2016; 354: 53–68.10.1016/j.wear.2016.02.018Search in Google Scholar
Arirachakaran S, Oglesby K, Malinowsky M, Shoham O, Brill J. An analysis of oil/water flow phenomena in horizontal pipes. In: SPE Production Operations Symposium, Society of Petroleum Engineers, 1989.10.2523/18836-MSSearch in Google Scholar
Ayello F, Robbins W, Richter S, Nesic S. Crude oil chemistry effects on inhibition of corrosion and phase wetting. Corrosion 2011. Paper No. 3149.10.5006/C2011-11060Search in Google Scholar
Bannwart A, Rodriguez O, Trevisan F, Vieira F, De Carvalho C. Experimental investigation on liquid-liquid-gas flow: flow patterns and pressure-gradient. J Petrol Sci Eng 2009; 65: 1–13.10.1016/j.petrol.2008.12.014Search in Google Scholar
Barker RJ, Hu X, Neville A, Cushnaghan S. Empirical prediction of carbon-steel degradation rates on an offshore oil and gas facility: predicting CO2 erosion-corrosion pipeline failures before they occur. SPE J 2014; 19: 425–436.10.2118/163143-PASearch in Google Scholar
Barnea D. A unified model for predicting flow-pattern transitions for the whole range of pipe inclinations. Int J Multiphase Flow 1987; 13: 1–12.10.1016/0301-9322(87)90002-4Search in Google Scholar
Beech IB, Sunner J. Biocorrosion: towards understanding interactions between biofilms and metals. Curr Opin Biotechnol 2004; 15: 181–186.10.1016/j.copbio.2004.05.001Search in Google Scholar PubMed
Brauner N. The prediction of dispersed flows boundaries in liquid-liquid and gas-liquid systems. Int J Multiphase Flow 2001; 27: 885–910.10.1016/S0301-9322(00)00056-2Search in Google Scholar
Brauner N, Ullmann A. Modeling of phase inversion phenomenon in two-phase pipe flows. Int J Multiphase Flow 2002; 28: 1177–1204.10.1016/S0301-9322(02)00017-4Search in Google Scholar
Bushnell L, Haas H. The utilization of certain hydrocarbons by microorganisms. J Bacteriol 1941; 41: 653.10.1128/jb.41.5.653-673.1941Search in Google Scholar PubMed PubMed Central
Cai J, Nesic S, De Waard C. Modeling of water wetting in oil-water pipe flow. In: Corrosion 2004, NACE International.10.5006/C2004-04663Search in Google Scholar
Cai J, Li C, Tang X, Ayello F, Richter S, Nesic S. Experimental study of water wetting in oil-water two phase flow – horizontal flow of model oil. Chem Eng Sci 2012; 73: 334–344.10.1016/j.ces.2012.01.014Search in Google Scholar
Chen X. Application of computational fluid dynamics (CFD) to flow simulation and erosion prediction in single-phase and multiphase flow. PhD thesis, University of Tulsa, 2004.Search in Google Scholar
Chen DS, Zhou YZ, Liu M, Guo KW, Wei WJ. Studies on corrosion behavior of Q235 steel by iron bacteria, sulfate-reducing bacteria and total general bacteria in sedimentary water of storage tank. Adv Mater Res Trans Tech Publ 2011; 337: 281–284.10.4028/www.scientific.net/AMR.337.281Search in Google Scholar
Chen S, Liu K, Liu C, Wang D, Ba D, Xie Y, Du G, Ba Y, Lin Q. Effects of surface tension and viscosity on the forming and transferring process of microscale droplets. Appl Surf Sci 2016; 388: 196–202.10.1016/j.apsusc.2016.01.205Search in Google Scholar
Craig B. Predicting the conductivity of water-in-oil solutions as a means to estimate corrosiveness. Corrosion 1998; 54: 657–662.10.5006/1.3287645Search in Google Scholar
Davies D, Burstein G. The effects of bicarbonate on the corrosion and passivation of iron. Corrosion 1980; 36: 416–422.10.5006/0010-9312-36.8.416Search in Google Scholar
De Waard C, Lotz U. Prediction of carbon dioxide corrosion of carbon steel. Corrosion 1993.10.5006/C1993-93069Search in Google Scholar
De Waard C, Milliams D. Carbonic acid corrosion of steel. Corrosion 1975; 31: 177–181.10.5006/0010-9312-31.5.177Search in Google Scholar
Dong F, Jiang Z, Qiao X, Xu L. Application of electrical resistance tomography to two-phase pipe flow parameters measurement. Flow Meas Instrum 2003; 14: 183–192.10.1016/S0955-5986(03)00024-4Search in Google Scholar
Eckert R. Emphasis on biofilms can improve mitigation of microbiologically influenced corrosion in oil and gas industry. Corros Eng Sci Technol 2015; 50: 163–168.10.1179/1743278214Y.0000000248Search in Google Scholar
Efird KD, Smith JL, Davis N, Blevins S. The crude oil effect on steel corrosion: wetability preference and brine chemistry. In: Corrosion 2004, NACE International.10.5006/C2004-04366Search in Google Scholar
El-hoshoudy A, Farag A, Ali O, El-Batanoney M, Desouky S, Ramzi M. New correlations for prediction of viscosity and density of Egyptian oil reservoirs. Fuel 2013; 112: 277–282.10.1016/j.fuel.2013.05.045Search in Google Scholar
Eliyan FF, Mahdi E-S, Alfantazi A. Electrochemical evaluation of the corrosion behaviour of API-X100 pipeline steel in aerated bicarbonate solutions. Corros Sci 2012; 58: 181–191.10.1016/j.corsci.2012.01.015Search in Google Scholar
Fingas M, Fieldhouse B. Formation of water-in-oil emulsions and application to oil spill modelling. J Hazard Mater 2004; 107: 37–50.10.1016/j.jhazmat.2003.11.008Search in Google Scholar PubMed
Flores J. Oil-water flow in vertical and deviated wells. Ph.D. thesis, University of Tulsa, 1997.Search in Google Scholar
Gandhi B, Borse S. Effects of particle size and size distribution on estimating erosion wear of cast iron in sand-water slurries. IJEMS 2002; 9: 480–486.Search in Google Scholar
Gaweł B, Lesaint C, Bandyopadhyay S, Øye G. Role of physicochemical and interfacial properties on the binary coalescence of crude oil drops in synthetic produced water. Energy Fuels 2015; 29: 512–519.10.1021/ef501847qSearch in Google Scholar
Gaweł B, Nourani M, Tichelkamp T, Øye G. Influence of the wettability of particles on the morphology and stability of crude oil-particle aggregates in synthetic produced water. J Petrol Sci Eng 2016; 139: 198–204.10.1016/j.petrol.2015.12.019Search in Google Scholar
Gece G. The use of quantum chemical methods in corrosion inhibitor studies. Corros Sci 2008; 50: 2981–2992.10.1016/j.corsci.2008.08.043Search in Google Scholar
Gonzalez J, Ramirez R, Hallen J, Guzman R. Hydrogen-induced crack growth rate in steel plates exposed to sour environments. Corrosion 1997; 53: 935–943.10.5006/1.3290278Search in Google Scholar
Guan SX, Ye HC, Wang XX, Fan YY. Research on detecting to bacterial concentration by cultivation-microscopy method in the oil field sewage. Adv Mater Res Trans Tech Publ 2013; 2837–2844.10.4028/www.scientific.net/AMR.726-731.2837Search in Google Scholar
Guan F, Zhai X, Duan J, Zhang J, Li K, Hou B. Influence of sulfate-reducing bacteria on the corrosion behavior of 5052 aluminum alloy. Surf Coat Technol 2017; 316: 171–179.10.1016/j.surfcoat.2017.02.057Search in Google Scholar
Gulbrandsen E, Nesic S, Stangeland A, Burchardt T, Sundfaer B, Hesjevik S, Skjerve S. Effect of pre-corrosion on the performance of inhibitors for CO2 corrosion of carbon steel. In: Corrosion National Association of Corrosion Engineers Annual Conference, NACE, 1998.10.5006/C1998-98013Search in Google Scholar
Hajirezaie S, Pajouhandeh A, Hemmati-Sarapardeh A, Pournik M, Dabir B. Development of a robust model for prediction of under-saturated reservoir oil viscosity. J Mol Liquids 2017; 229: 89–97.10.1016/j.molliq.2016.11.088Search in Google Scholar
Harrigan WF, McCance ME. Laboratory methods in food and dairy microbiology. London: Academic Press, Inc., 1976.Search in Google Scholar
Hemmati-Sarapardeh A, Shokrollahi A, Tatar A, Gharagheizi F, Mohammadi AH, Naseri A. Reservoir oil viscosity determination using a rigorous approach. Fuel 2014; 116: 39–48.10.1016/j.fuel.2013.07.072Search in Google Scholar
Hernandez S, Duplat S, Vera JR, Baron E. A statistical approach for analyzing the inhibiting effect of different types of crude oil in CO2 corrosion of carbon steel. In: Corrosion 2002, NACE International.10.5006/C2002-02293Search in Google Scholar
Hernandez S, Nešić S, Weckman G, Ghai V. Use of artificial neural networks for predicting crude oil effect on carbon dioxide corrosion of carbon steels. Corrosion 2006; 62: 467–482.10.5006/1.3279905Search in Google Scholar
Hinze J. Fundamentals of the hydrodynamic mechanism of splitting in dispersion processes. AIChE J 1955; 1: 289–295.10.1002/aic.690010303Search in Google Scholar
Horváth-Szabó G, Masliyah JH, Elliott JA, Yarranton HW, Czarnecki J. Adsorption isotherms of associating asphaltenes at oil/water interfaces based on the dependence of interfacial tension on solvent activity. J Colloid Interface Sci 2005; 283: 5–17.10.1016/j.jcis.2004.08.174Search in Google Scholar PubMed
Huang YH, Zhang TC. Effects of dissolved oxygen on formation of corrosion products and concomitant oxygen and nitrate reduction in zero-valent iron systems with or without aqueous Fe2+. Water Res 2005; 39: 1751–1760.10.1016/j.watres.2005.03.002Search in Google Scholar PubMed
Huang X, Qi Y, Chen C, Yu H, Lu G. Effect of environmental factors on corrosion behaviour of L360QCS pipeline steel in H2S/CO2 environments. Corros Eng Sci Technol 2015; 50: 169–177.10.1179/1743278215Y.0000000011Search in Google Scholar
Ingham B, Ko M, Laycock N, Burnell J, Kappen P, Kimpton J, Williams D. In situ synchrotron X-ray diffraction study of scale formation during CO2 corrosion of carbon steel in sodium and magnesium chloride solutions. Corros Sci 2012; 56: 96–104.10.1016/j.corsci.2011.11.017Search in Google Scholar
Javed M, Stoddart P, Wade S. Corrosion of carbon steel by sulphate reducing bacteria: initial attachment and the role of ferrous ions. Corros Sci 2015; 93: 48–57.10.1016/j.corsci.2015.01.006Search in Google Scholar
Jiang X, Nešić S, Kinsella B, Brown B, Young D. Electrochemical investigation of the role of Cl− on localized carbon dioxide corrosion behavior of mild steel. Corrosion 2012; 69: 15–24.10.5006/0620Search in Google Scholar
Jin H, Chen X, Zheng Z, Ou G, Liu W. Failure analysis of multiphase flow corrosion-erosion with three-way injecting water pipe. Eng Fail Anal 2017; 73: 46–56.10.1016/j.engfailanal.2016.12.005Search in Google Scholar
Karabelas A. Vertical distribution of dilute suspensions in turbulent pipe flow. AIChE J 1977; 23: 426–434.10.1002/aic.690230404Search in Google Scholar
Karabelas A. Droplet size spectra generated in turbulent pipe flow of dilute liquid/liquid dispersions. AIChE J 1978; 24: 170–180.10.1002/aic.690240203Search in Google Scholar
Kee KE. A study of flow patterns and surface wetting in gas-oil-water flow. Ph.D. thesis, Ohio University, 2014.Search in Google Scholar
Kee KE, Babic M, Richter S, Paolinelli L, Li W, Nesic S. Flow patterns and water wetting in gas-oil-water three-phase flow – a flow loop study. Corrosion 2015. Paper No. 6113.10.5006/C2015-06113Search in Google Scholar
Kermani M, Morshed A. Carbon dioxide corrosion in oil and gas production – a compendium. Corrosion 2003; 59: 659–683.10.5006/1.3277596Search in Google Scholar
Kokal S. Crude oil emulsions: a state-of-the-art review. In: SPE Annual Technical Conference and Exhibition, Society of Petroleum Engineers, 2002.10.2118/77497-MSSearch in Google Scholar
Kouba GE. Mechanistic models for droplet formation and breakup. In: ASME/JSME 2003 4th Joint Fluids Summer Engineering Conference, American Society of Mechanical Engineers, 2003.Search in Google Scholar
Koyama M, Akiyama E, Tsuzaki K. Hydrogen embrittlement in a Fe-Mn-C ternary twinning-induced plasticity steel. Corros Sci 2012; 54: 1–4.10.1016/j.corsci.2011.09.022Search in Google Scholar
Kubie J, Gardner G. Drop sizes and drop dispersion in straight horizontal tubes and in helical coils. Chem Eng Sci 1977; 32: 195–202.10.1016/0009-2509(77)80105-XSearch in Google Scholar
Kumara W, Halvorsen B, Melaaen MC. Pressure drop, flow pattern and local water volume fraction measurements of oil-water flow in pipes. Meas Sci Technol 2009; 20: 114004.10.1088/0957-0233/20/11/114004Search in Google Scholar
Lang X-f, Qiu L-n, Gong A-j, Ma X. Review on microbiologically influenced corrosion and antisepsis techniques. Total Corrosion Control 2009; 10: 011.Search in Google Scholar
Larsen KR. Managing corrosion of pipelines that transport crude oils. Mater Perform 2013; 52: 28–35.10.5006/MP2013_52_9-28Search in Google Scholar
Lee W, Lewandowski Z, Morrison M, Characklis WG, Avci R, Nielsen PH. Corrosion of mild steel underneath aerobic biofilms containing sulfate-reducing bacteria. Part II: at high dissolved oxygen concentration. Biofouling 1993; 7: 217–239.10.1080/08927019309386255Search in Google Scholar
Lévesque J, Hermawan H, Dubé D, Mantovani D. Design of a pseudo-physiological test bench specific to the development of biodegradable metallic biomaterials. Acta Biomater 2008; 4: 284–295.10.1016/j.actbio.2007.09.012Search in Google Scholar PubMed
Levy AV, Chik P. The effects of erodent composition and shape on the erosion of steel. Wear 1983; 89: 151–162.10.1016/0043-1648(83)90240-5Search in Google Scholar
Li C, Tang X, Ayello F, Cai J, Nesic S, Cruz CIT, Al-Khamis JN. Experimental study on water wetting and CO2 corrosion in oil-water two-phase flow. In: NACE Corrosion, 2006: 6595.10.5006/C2006-06595Search in Google Scholar
Lin R, Song D, Zhou G, Wang X, Tian X, Yang K. Hydrogen sulfide formation mechanism in the process of thermal recovery. Acta Petrol Sin 2014; 35: 1153–1159.Search in Google Scholar
Liu H, Fu C, Gu T, Zhang G, Lv Y, Wang H, Liu H. Corrosion behavior of carbon steel in the presence of sulfate reducing bacteria and iron oxidizing bacteria cultured in oilfield produced water. Corros Sci 2015; 100: 484–495.10.1016/j.corsci.2015.08.023Search in Google Scholar
Lopes F, Morin P, Oliveira R, Melo L. The influence of nickel on the adhesion ability of Desulfovibrio desulfuricans. Colloids Surf B Biointerfaces 2005; 46: 127–133.10.1016/j.colsurfb.2005.07.020Search in Google Scholar PubMed
Lopes F, Morin P, Oliveira R, Melo L. Interaction of Desulfovibrio desulfuricans biofilms with stainless steel surface and its impact on bacterial metabolism. J Appl Microbiol 2006; 101: 1087–1095.10.1111/j.1365-2672.2006.03001.xSearch in Google Scholar PubMed
Lotz U, Van Bodegom L, Ouwehand C. The effect of type of oil or gas condensate on carbonic acid corrosion. Corrosion 1991; 47: 635–645.10.5006/1.3585301Search in Google Scholar
Maruthamuthu S, Kumar BD, Ramachandran S, Anandkumar B, Palanichamy S, Chandrasekaran M, Subramanian P, Palaniswamy N. Microbial corrosion in petroleum product transporting pipelines. Ind Eng Chem Res 2011; 50: 8006–8015.10.1021/ie1023707Search in Google Scholar
Mendez C, Dupla S, Hernandez S, Vera JR. On the mechanism of corrosion inhibition by crude oils. In: Corrosion 2001, NACE International.10.5006/C2001-01044Search in Google Scholar
Meresht ES, Farahani TS, Neshati J. Failure analysis of stress corrosion cracking occurred in a gas transmission steel pipeline. Eng Fail Anal 2011; 18: 963–970.10.1016/j.engfailanal.2010.11.014Search in Google Scholar
Ming LI, Xiaogang LI, Chen G. Influencing factors of hydrogen diffusion in hydrogen sulfide environment. J Univ Sci Technol Beijing 2007; 29: 39–44.Search in Google Scholar
NACE. Field monitoring of bacterial growth in oil and gas systems. Houston, TX: NACE, 2004.Search in Google Scholar
Nädler M, Mewes D. Flow induced emulsification in the flow of two immiscible liquids in horizontal pipes. Int J Multiphase Flow 1997; 23: 55–68.10.1016/S0301-9322(96)00055-9Search in Google Scholar
Nešić S. Key issues related to modelling of internal corrosion of oil and gas pipelines – a review. Corros Sci 2007; 49: 4308–4338.10.1016/j.corsci.2007.06.006Search in Google Scholar
Nesic S, Postlethwaite J. Relationship between the structure of disturbed flow and erosion-corrosion. Corrosion 1990; 46: 874–880.10.5006/1.3580852Search in Google Scholar
Nesic S, Sun W. Corrosion in acid gas solutions. In: Shreir’s corrosion, vol. 2, 2010: 1270–1298.10.1016/B978-044452787-5.00055-XSearch in Google Scholar
Nešić S, Wang S, Fang H, Sun W, Lee J. A new updated model of CO2/H2S corrosion in multiphase flow. In: Corrosion 2008 Conference and Expo, NACE International, New Orleans.10.5006/C2008-08535Search in Google Scholar
Nyborg R. Initiation and growth of mesa corrosion attack during CO2 corrosion of carbon steel. Houston, TX: NACE International, 1998.10.5006/C1998-98048Search in Google Scholar
Paolinelli LD, Nesic S. Hydrodynamic and phase wetting criteria to assess corrosion risk in two-phase oil-water pipe flow. In: Corrosion 2016, NACE International.10.5006/C2016-07408Search in Google Scholar
Papavinasam S, Doiron A, Panneerselvam T, Revie R. Effect of hydrocarbons on the internal corrosion of oil and gas pipelines. Corrosion 2007; 63: 704–712.10.5006/1.3278419Search in Google Scholar
Pawar AB, Caggioni M, Ergun R, Hartel RW, Spicer PT. Arrested coalescence in pickering emulsions. Soft Matter 2011; 7: 7710–7716.10.1039/c1sm05457kSearch in Google Scholar
Perazzo A, Preziosi V, Guido S. Phase inversion emulsification: current understanding and applications. Adv Colloid Interface Sci 2015; 222: 581–599.10.1016/j.cis.2015.01.001Search in Google Scholar PubMed
Ping X, Jin W, Yajun Z, Zhaoyi X, Xiaodong L, Ting L. Comparison of microbiologically induced corrosion on metals in industrial recycling cooling system makeup by municipal reclaimed water. J Tianjin Univ 2013; 2: 007.Search in Google Scholar
Pots BF, Kapusta SD, John RC, Thomas M, Rippon IJ, Whitham T, Girgis M. Improvements on De Waard-Milliams corrosion prediction and applications to corrosion management. In: Corrosion 2002, NACE International.10.5006/C2002-02235Search in Google Scholar
Pots BF, Hendriksen EL, Hollenberg J. What are the real influences of flow on corrosion? In: Corrosion 2006, NACE International.10.5006/C2006-06591Search in Google Scholar
Pouraria H, Seo JK, Paik JK. A numerical study on water wetting associated with the internal corrosion of oil pipelines. Ocean Eng 2016; 122: 105–117.10.1016/j.oceaneng.2016.06.022Search in Google Scholar
Raja PB, Sethuraman MG. Natural products as corrosion inhibitor for metals in corrosive media – a review. Mater Lett 2008; 62: 113–116.10.1016/j.matlet.2007.04.079Search in Google Scholar
Ramanarayanan T, Smith S. Corrosion of iron in gaseous environments and in gas-saturated aqueous environments. Corrosion 1990; 46: 66–74.10.5006/1.3585068Search in Google Scholar
Richter S, Babic M, Tang X, Robbins W, Nesic S. Categorization of crude oils based on their ability to inhibit corrosion and alter the steel wettability. In: NACE Corrosion, San Antonio, 2014: 4247.10.5006/C2014-4247Search in Google Scholar
Russell T, Hodgson G, Govier G. Horizontal pipeline flow of mixtures of oil and water. Can J Chem Eng 1959; 37: 9–17.10.1002/cjce.5450370104Search in Google Scholar
Ruzic V, Veidt M, Nešić S. Protective iron carbonate films – Part 1: mechanical removal in single-phase aqueous flow. Corrosion 2006; 62: 419–432.10.5006/1.3278279Search in Google Scholar
Sandor M, Cheng Y, Chen S. Improved correlations for heavy-oil viscosity prediction with NMR. J Petrol Sci Eng 2016; 147: 416–426.10.1016/j.petrol.2016.09.004Search in Google Scholar
Schmitt G, Stradmann N. Wettability of steel surfaces at CO2 corrosion conditions. 1. Effect of surface active compounds in aqueous and hydrocarbon media. Houston, TX: NACE International, 1998.10.5006/C1998-98028Search in Google Scholar
Segev A. Mechanistic model for estimating water dispersion in crude oil flow. New York, NY: American Institute of Chemical Engineers, 1985.Search in Google Scholar
Selker A, Sleicher C. Factors affecting which phase will disperse when immiscible liquids are stirred together. Can J Chem Eng 1965; 43: 298–301.10.1002/cjce.5450430606Search in Google Scholar
Sharifi M, Young B. Electrical resistance tomography (ERT) applications to chemical engineering. Chem Eng Res Des 2013; 91: 1625–1645.10.1016/j.cherd.2013.05.026Search in Google Scholar
Sheng J. Modern chemical enhanced oil recovery: theory and practice. Houston, Texas, USA: Gulf Professional Publishing, 2010.Search in Google Scholar
Shoham O. Mechanistic modeling of gas-liquid two-phase flow in pipes. Richardson, TX: Society of Petroleum Engineers, 2006.10.2118/9781555631079Search in Google Scholar
Shoham O, Arirachakaran S, Brill JP. Two-phase flow splitting in a horizontal reduced pipe tee. Chem Eng Sci 1989; 44: 2388–2391.10.1016/0009-2509(89)85174-7Search in Google Scholar
Simoni L, Caselani JQ, Ramos LB, Schroeder RM, de Fraga Malfatti C. The influence of calcareous deposits on hydrogen uptake and embrittlement of API 5CT P110 steel. Corros Sci 2017; 118: 178–189.10.1016/j.corsci.2017.02.007Search in Google Scholar
Smart JS. Wettability: a major factor in oil and gas system corrosion. Mater Perform 2001; 40: 54–59.Search in Google Scholar
Smith L, De Waard K, Craig BD. The influence of crude oils on well tubing corrosion rates. In: Corrosion 2003, NACE International.10.5006/C2003-03629Search in Google Scholar
Song X, Yang Y, Yu D, Lan G, Wang Z, Mou X. Studies on the impact of fluid flow on the microbial corrosion behavior of product oil pipelines. J Petrol Sci Eng 2016; 146: 803–812.10.1016/j.petrol.2016.07.035Search in Google Scholar
Srdjan N, Cai J, Lee KJ. A multiphase flow and internal corrosion prediction model for mild steel pipeline. In: Corrosion 2005, NACE International.Search in Google Scholar
Stroe M, Passade-Boupat N, Bonis M, Adams B. Inhibitive properties of crude oils: can we count on them. In: Corrosion 2011, NACE International.10.5006/C2011-11061Search in Google Scholar
Taitel Y, Dukler A. A model for predicting flow regime transitions in horizontal and near horizontal gas-liquid flow. AIChE J 1976; 22: 47–55.10.1002/aic.690220105Search in Google Scholar
Tang X. Effect of surface state on water wetting and carbon dioxide corrosion in oil-water two-phase flow. Ohio University, 2011.Search in Google Scholar
Thaker J, Banerjee J. Influence of intermittent flow sub-patterns on erosion-corrosion in horizontal pipe. J Petrol Sci Eng 2016; 145: 298–320.10.1016/j.petrol.2016.05.006Search in Google Scholar
Trallero JL. Oil-water flow patterns in horizontal pipes. Ph.D. thesis, University of Tulsa, Tulsa, OK, 1995.Search in Google Scholar
Tsahalis D. Conditions for the entrainment of settled water in crude oil and product pipelines. In: 83rd Annual AIChE Meeting, March 1977.Search in Google Scholar
Urquidi-Macdonald M, Tewari A, Ayala HLF. A neuro-fuzzy knowledge-based model for the risk assessment of microbiologically influenced corrosion in crude oil pipelines. Corrosion 2014; 70: 1157–1166.10.5006/1174Search in Google Scholar
Vielma MA, Atmaca S, Sarica C, Zhang H-Q. Characterization of oil/water flows in horizontal pipes. SPE Proj Facil Constr 2008; 3: 1–21.10.2118/109591-PASearch in Google Scholar
Vigneaux P, Chenais P, Hulin J. Liquid-liquid flows in an inclined pipe. AIChE J 1988; 34: 781–789.10.1002/aic.690340508Search in Google Scholar
Wan W, Xiong J, Li Y, Tang Q, Liang M. Erosion-corrosion behavior of Ti (C, N)-based cermets containing different secondary carbides. Int J Refract Metals Hard Mater 2017; 66: 180–187.10.1016/j.ijrmhm.2017.03.018Search in Google Scholar
Wang ZM, Zhang J. Corrosion of multiphase flow pipelines: the impact of crude oil. Corros Rev 2016; 34: 17–40.10.1515/corrrev-2015-0053Search in Google Scholar
Wang J, Giridharan V, Shanov V, Xu Z, Collins B, White L, Jang Y, Sankar J, Huang N, Yun Y. Flow-induced corrosion behavior of absorbable magnesium-based stents. Acta Biomater 2014a; 10: 5213–5223.10.1016/j.actbio.2014.08.034Search in Google Scholar PubMed
Wang Z, Han X, Zhang J, Wang Z. In situ observation of CO2 corrosion under high pressure. Corros Eng Sci Technol 2014b; 49: 352–356.10.1179/1743278213Y.0000000144Search in Google Scholar
Wang J, Jang Y, Wan G, Giridharan V, Song G-L, Xu Z, Koo Y, Qi P, Sankar J, Huang N. Flow-induced corrosion of absorbable magnesium alloy: in-situ and real-time electrochemical study. Corros Sci 2016; 104: 277–289.10.1016/j.corsci.2015.12.020Search in Google Scholar PubMed PubMed Central
Wegmann A, Melke J, von Rohr PR. Three phase liquid-liquid-gas flows in 5.6 mm and 7mm inner diameter pipes. Int J Multiphase Flow 2007; 33: 484–497.10.1016/j.ijmultiphaseflow.2006.10.004Search in Google Scholar
Wicks M, Fraser J. Entrainment of water by flowing oil. Mater Perform 1975; 14: 9–12.Search in Google Scholar
Xia Z, Chou K, Szklarska-Smialowska Z. Pitting corrosion of carbon steel in CO2-containing NaCl brine. Corrosion 1989; 45: 636–642.10.5006/1.3579317Search in Google Scholar
Xu X-X. Study on oil-water two-phase flow in horizontal pipelines. J Petrol Sci Eng 2007; 59: 43–58.10.1016/j.petrol.2007.03.002Search in Google Scholar
Xu J-y, Li D-h, Guo J, Wu Y-x. Investigations of phase inversion and frictional pressure gradients in upward and downward oil-water flow in vertical pipes. Int J Multiphase Flow 2010; 36: 930–939.10.1016/j.ijmultiphaseflow.2010.08.007Search in Google Scholar
Xu G-l, Zhang G-z, Liu G, Ullmann A, Brauner N. Trapped water displacement from low sections of oil pipelines. Int J Multiphase Flow 2011; 37: 1–11.10.1016/j.ijmultiphaseflow.2010.09.003Search in Google Scholar
Xu G, Cai L, Ullmann A, Brauner N. Trapped water flushed by flowing oil in upward-inclined oil pipelines. In: 9th International Pipeline Conference, American Society of Mechanical Engineers, 2012.10.1115/IPC2012-90680Search in Google Scholar
Yang H, Xue XP, Fu Z-X, Wang W, Song K, Huyan T, Wang H. Research advances on microbiologically influenced corrosion and its prevention measures in marine environment. Chem Bioeng 2010; 1: 002.Search in Google Scholar
Yang S, Richter S, Robbins W, Nesic S. Evaluation of the protectiveness of a paraffin layer in CO2 corrosion of mild steel. In: Corrosion 2012, NACE International.10.5006/C2012-01323Search in Google Scholar
Yiing-Mei W. Entrainment method enhanced to account for oil’s water content. Int J Multiphase Flow 1996; 1001: 138.10.1016/S0301-9322(97)88500-XSearch in Google Scholar
Young T. An essay on the cohesion of fluids. Philos Trans R Soc Lond 1805; 95: 65–87.10.1098/rstl.1805.0005Search in Google Scholar
Yusuf N, Al-Wahaibi Y, Al-Wahaibi T, Al-Ajmi A, Olawale A, Mohammed I. Effect of oil viscosity on the flow structure and pressure gradient in horizontal oil-water flow. Chem Eng Res Des 2012; 90: 1019–1030.10.1016/j.cherd.2011.11.013Search in Google Scholar
Zhang J, Wang ZL, Wang ZM, Han X. Chemical analysis of the initial corrosion layer on pipeline steels in simulated CO2-enhanced oil recovery brines. Corros Sci 2012; 65: 397–404.10.1016/j.corsci.2012.08.045Search in Google Scholar
Zhang R, Yang L, Tu R, Huo J, Wang J, Zhou J, Chen D. Emulsion phase inversion from oil-in-water (1) to water-in-oil to oil-in-water (2) induced by in situ surface activation of CaCO3 nanoparticles via adsorption of sodium stearate. Colloids Surf A Physicochem Eng Aspects 2015; 477: 55–62.10.1016/j.colsurfa.2015.03.043Search in Google Scholar
Zhang N, Zeng D, Xiao G, Shang J, Liu Y, Long D, He Q, Singh A. Effect of Cl− accumulation on corrosion behavior of steels in H2S/CO2 methyldiethanolamine (MDEA) gas sweetening aqueous solution. J Nat Gas Sci Eng 2016; 30: 444–454.10.1016/j.jngse.2016.02.055Search in Google Scholar
Zhang J, Yuan H, Zhao J, Mei N. Viscosity estimation and component identification for an oil-water emulsion with the inversion method. Appl Thermal Eng 2017; 111: 759–767.10.1016/j.applthermaleng.2016.09.153Search in Google Scholar
Zhao W, Wang C, Zhang T, Yang M, Han B, Neville A. Effects of laser surface melting on erosion-corrosion of X65 steel in liquid-solid jet impingement conditions. Wear 2016; 362: 39–52.10.1016/j.wear.2016.05.006Search in Google Scholar
Zheng D, Che D, Liu Y. Experimental investigation on gas-liquid two-phase slug flow enhanced carbon dioxide corrosion in vertical upward pipeline. Corros Sci 2008; 50: 3005–3020.10.1016/j.corsci.2008.08.006Search in Google Scholar
©2017 Walter de Gruyter GmbH, Berlin/Boston
Articles in the same Issue
- Frontmatter
- In this issue
- Reviews
- Selected Cr(VI) replacement options for aluminum alloys: a literature survey
- Actual trends in the elaboration of advanced multifunctional coating systems for the efficient protection of lightweight aircraft alloys
- A review on stray current-induced steel corrosion in infrastructure
- A review of internal corrosion mechanism and experimental study for pipelines based on multiphase flow
- Original articles
- Effect of interfacial film on the corrosion behaviour of X80 pipeline steel in a neutral soil environment containing sulphate-reducing bacteria
- Corrosion behavior of new type Co-based superalloys with different Ni contents
- Synergistic inhibition effect of imidazoline derivative and potassium iodide on carbon steel corrosion in amidosulfuric acid solution
- Characterization of the Ag43Cu37Zn20 alloy surface after potentiostatic polarization using LDI-TOF mass spectrometry
Articles in the same Issue
- Frontmatter
- In this issue
- Reviews
- Selected Cr(VI) replacement options for aluminum alloys: a literature survey
- Actual trends in the elaboration of advanced multifunctional coating systems for the efficient protection of lightweight aircraft alloys
- A review on stray current-induced steel corrosion in infrastructure
- A review of internal corrosion mechanism and experimental study for pipelines based on multiphase flow
- Original articles
- Effect of interfacial film on the corrosion behaviour of X80 pipeline steel in a neutral soil environment containing sulphate-reducing bacteria
- Corrosion behavior of new type Co-based superalloys with different Ni contents
- Synergistic inhibition effect of imidazoline derivative and potassium iodide on carbon steel corrosion in amidosulfuric acid solution
- Characterization of the Ag43Cu37Zn20 alloy surface after potentiostatic polarization using LDI-TOF mass spectrometry
