Factors Governing the Failure of Subsea Critical Connector Bolts

Impressed current cathodic protection (ICCP) systems are commonly used to shield offshore drilling rigs, pipelines, and subsea equipment in the oil and gas industry. In underwater service conditions, water temperature, salinity and velocity play a major role in the longevity of subsea applications. Interactions between the preceding factors can induce catastrophic failure to critical systems while the underlying cause is unclear. This paper proposes an approach for elucidating the corrosion process accompanying underwater applications. The service conditions of underwater application are simulated in a multidisciplinary system that records various parameters such as water temperature, reference-electrode potential, and electric current at five-minute intervals during the 21 d of the experiment. This novel, experimental, and inexpensive ICCP system was developed on an “Arduino” microcontroller and applied to an actual ASTM A193 B7 bolt tightened on an ASTM A105 flange at different torque levels. Experimental results indicate a direct relationship between the water day-night temperature profile and the cathodic protection performance. Specifically, the ICCP performance declines with increasing temperature. When the ICCP system was activated, gas bubbles are generated on the metal surface. Presumably these bubbles could induce hydrogen embrittlement cracks which were observed in scanning electron microscope images of the bolt cross-sections.

Green technologies which help satisfy the growing demand for energy are expected to mature only over the next two to three decades.In the interim, the demand for natural gas is expected to rise by almost by 50%, by 2040, and oil consumption will continue to grow.Therefore, offshore hydrocarbon resources will continue to attract strong interest in the foreseeable future. 1 Offshore structures are often exposed to extreme marine conditions battered by waves and storms in a saline environment.At such, offshore rigs and subsea equipment require effective and affordable protection.Cathodic protection is widely used to safeguard marine structural steel, but over-protection during the application of impressed current cathodic protection (ICCP) can foster hydrogen embrittlement of the high-strength steel used in marine structures. 2 ICCP converts the anodic metallic node to the cathode of an electrochemical cell. 3The anodic and the cathodic reactions expresed by Eq. 1 and Eq. 2 respectivetly, refer to the corrosion reactions manifesting at the steel surface: Notably, the cathodic corrosion reaction given by Eq. 2 pertains to the oxygen reduction reaction (ORR). 4In ICCP, electrons with an impress current are supplied to the surface of the protected configuration, converting the area surface into a cathode.Considering the underwater conditions with a pH ranging between 7.8 to 8.2, and the effect of the ICCP on the voltage potential range that should be between 0.8 V and 1.1 V in the Pourbaix diagram, the dominant cathodic reaction is captured by the hydrogen evolution reaction (HER) expressed by in Eq. 3. The anodes used in an ICCP system are usually insoluble.Therefore, the anodic reaction accompanies the evolution of oxygen evolution as chemically given by Eq. 4: 5 Cathodic protection of marine structures is a complex problem requiring tailored made solutions for each application.To achieve the necessary level of protection, one must consider the various parameters governing the performance of the cathodic protection, such as the seawater temperature, the water salinity, the water flow velocity, and the magnitude of protective current. 6Chief among the main factors controlling long-term marine corrosion comprise the exposure time, temperature, salinity, and water velocity. 7Time is the main parameter affecting the underwater corrosion processes, but in most studies, cathodic protection is simulated over periods of a few hours.To better understand the impact of the ICCP on underwater corrosion processes, an extended experimentation time is required.
As expected, the water temperature also affects the performance of the cathodic protection system.The electric current density diminishes more at 28 °C (tropical conditions) than at 5 °C (arctic conditions).This phenomenon is related to the formation of deposits on the surface of the steel. 8Seawater temperature dominantly influences the electric current density and, thus, the properties of the final deposited layer. 9,10A rise in water temperature causes an immediate surge in current density.Water decomposition (reduction (is more intense at higher temperatures and the growth of calcareous deposits reduces the ORR on the metal surface. Seawater velocity can affect the corrosion rate of metals in many ways.First, it replenishes the amount of oxygen reaching the metal's surface and accelerates the deterioration of protective films.Boosting the seawater velocity from static to 1 m s −1 may double the corrosion rate.The concentration of dissolved oxygen is a function not only of temperature but also of the magnitude of movement of the water (currents). 11The degree of water flow rate varies regionally and seasonably. 12ttempting to model the effects of the main environmental factors on the underwater corrosion process is a challenging task because of the complex combined effect of the marine environment.Overall, the corrosion reaction is mostly affected by the combined effect of water velocity, temperature, and salinity over the range of lab test conditions. 13The various parameters of marine conditions constantly and simultaneously interact with each other and hence influence the underwater process.The time of exposure to the environment is critical in cathodic protection. 14The collective influence of the main factors on an underwater corrosion process over time remains uncertain and merits more rigorous nvestigation.
In this study, a novel approach has been devised in the context of an experimental investigation.Simulating the working conditions of z E-mail: oferursa@gmail.comECS Advances, 2023 2 041501 real subsea critical connector bolts and monitoring the data of several parameters.More specifically, the parameters comprised the water temperature, subsea water currents, and most importantly the ICCP electrical system parameters.
Operationally, the various measured data from the ICCP system were collected by a functional system which included an ammeter which registered the current of the ICCP electrical circuit, and a thermometer which logged the water temperature in the aquarium.In addition, a reference electrode measured the electrical potential difference relative to the environment at six distinct locations.Temporally, each instrument obtained a measurement every 5 min over an experimental timeframe of 21 d while the readings were recorded digitally and automatically in an Excel spreadsheet.
Experimentally, the ICCP system was tested on a configuration that consisted of an ASTM A105 15 flange with eight ASTM A193 B7 16 bolts in an aquarium containing 3.5% of sodium chloride (NaCl) salt.As mentioned earlier, the test time stretched to 21 d.Bolted connections are an important component of subsea operations involving a subsea wellhead and often a marine riser system.Thousands of bolting components (e.g., threaded bolts, studs, and nuts) are used in a wide spectrum of applications. 17Based on the cross-sectional area of a bolt, the maximum allowable tensile preloading on the bolt ranged between 67% and 73% of the bolt's material yield point stress (Y.P.), which equals 724 MPa (API Spec 6 A 18 and ISO10432 19 ).Suffice to mention that when the bolt is subjected to a torque equal to or greater than 80% of the material's Y. P., the electrochemical activity the system is experiencing can dramatically surge. 20Moreover, the concentration of stress at the crack tip causes a local high plastic stress that accelerates the electrochemical activity.As a result the anodic reaction, the absorption and diffusion of the hydrogen increases, thus speeding-up the propagation of the cracks. 21CCP systems play a critical role in preserving the operational envelope of subsea equipment.Yet, few publications have reported the major ICCP systems parameters that affect the corrosion protection system's performance as well as the impact cathodic protection exerts on the subsea corrosion process.Collectively, the information gleaned from the experimental system and the analysis are intended to assist the prevailing state of understanding of subsea corrosion processes and to determine the relationships between the parameters governing this process.

Corrosion Experimental Set-Up
As illustrated in Fig. 1, the small-scale laboratory ICCP system consists of the following components: 1. Power switch 2. A DC power supply.
The microcontroller and the main computing unit was an Arduino Mega 2560, which supplied direct electric current to the ICCP system.Featuring the T-mega 2560 platform, the Arduino Mega 2560 is an electronic board.It exhibits 54 digital input/output connectors, a USB port, a power jack, and a reset button.Because the Arduino circuit board is based on open-source code, it can be conveniently programmed and operated at reasonable cost.Many examples and numerous models are freely available online.Arduino can supply up to five volts of power, which is adequate for the ICCP system.Arduino can also be programmed as an ampere meter.The voltage drop denoted by R across the resistor is quantified as the difference between the voltage across the two ends of the resistor, which could be measured.Utilizing Ohm's Law, the current flow across the device can then be determined.A 1 KΩ potentiometer permits adjustments in the system's electric current. 22eing immersed in water, subsea applications warrant an insoluble anode.Some of the important attributes of insoluble anodes comprise their mass density, electrical resistivity, energy consumption, and anode current density (A m −2 ). 5 To ensure the relevant parameters, the present study employs a platinum anode.
Mechanically, the configuration to be protected consists of an ASTM A105 flange and eight ASTM A193 B7 bolts subjected to three different levels of torques whose magnitudes comprised: (a) 451 Nm (Y.P.), (b) 341 Nm (75% of Y. P.), and 114 Nm (25% of Y. P.).The eight bolt-flange arrangements are listed in Table I and illustrated in color in Fig. 2.
All bolts were fastened at a specific moment using a special precision torque meter with a digital display.When the torque meter registered the desired torque magnitude, it sounded an alarm and flickered red lights.After the bolts were fastened to the flange, the configuration was immersed in the aquarium containing 3.5% NaCl solution for a duration of 21 d.ECS Advances, 2023 2 041501 Corrosion protection mainly concerns the flange, the bolt, and the nut set-up.Collectively the surface area of the bolt was 0.27 m 2 .In the Mediterranean Sea, the recommended current density amounts to 110 mA m −2 according to the standard EN-13173. 23A variable electrical resistor was used to control the electrical current (Iin) required in our configuration.The electrical potential difference between the environment and the protected surface of the flange-bolt system was read by an Ag/AgCl reference electrode.
Together with calibrating the electric current in the ICCP system it is also essential to examine the accuracy of the system.For this purpose, we utilized a multi-meter.Corrosion was efficiently arrested when the electrical potential as indicated by the reference electrode ranged between −0.7 V and −0.75 V Ag/AgCl.For the record, standard EN-12495 24 recommends more negative potentials ranging from −0.80 V to −1.10 V Ag/AgCl.A multi-meter was used to confirm that the supplied electrical current resided within the standard range.If necessary, the current was adjusted using a variable resistor until the reference electrode read −850 mV.
The rate of corrosion can be measured using the rate of mass loss per unit area exposed to the corrosive environment or the depth of corrosion penetration per year [mm year −1 Here Δm is the mass loss at time t of the structure, S is the exposed surface area, and ρ is the mass density of the material.To calculate the corrosion rate and determine the bolts mass loss during the 21 d of the experiment, the bolts were weighed before and after the experiment.Equation 5 was subsequently used to obtain the corrosion rate.Primarily the experiment aimed at ensuring the correct functioning of the ICCP system and the protection of the flange-bolt configuration.
The aquarium was equipped with a thermometer capable of monitoring the salty water temperature.Temperature readings were embedded in the Excel spreadsheet created by the Arduino microcontroller.A water pump for simulating the water current was also fittedto the experimental set-up in the water tank (Fig. 3).
After 21 d, the flange was retrieved from the aquarium (Fig. 4).The bolts were dismantled and cleaned with 200 g of sodium hydroxide (NaOH) and 20 g of Zinc chips and regent water for 1 l solution that heated to 80-90 for 40 min, according to ASTM G1    ECS Advances, 2023 2 041501 standard. 25Finally, the bolts were subjected to microscopic evaluation using a scanning electron microscope (SEM).

Results
Provided that the new ICCP system offers good protection, the corrosion rate of the exposed parts, namely, the flange and the bolts configuration will be kept at bay.ISO 9223, 26 "very good" corrosion rates are defined in the range of 0.02 to 0.1 mm of metal loss per annum.Results pertaining to the corrosion rates of the protected bolts of the present configuration ranged from 0.043 to 0.082 mm y −1 (Fig. 5).Hence, the level of protection offered by the ICCP unit can be classified as "very good." Analyzing the findings of this investigation it was possible to establish a correlation between the rate of the bolts in the ICCP system and the torque magnitude of the tightened bolts.Increasing the torque from 114 Nm (25% of Y. P.) to 341 Nm (75% of Y. P.), the corrosion rate surged from 0.044 to 0.072 mm y −1 or an equivalent rise of 63.5%.Under the application of the largest torque of 455 Nm which matches the steel's yield point, the corrosion rate accelerated to 0.082 mm y −1 translating into an 87% faster rate than when the bolt was tightened at 114 Nm (25% of the Y. P.).
Once the proper functioning of the ICCP system was verified, formal collection of the data commenced.During the 21 d of the experiment, the set-up retrieved more than 6,000 readings.Scrutinizing the data, attention was placed at identifying useful patterns, connections, and correlations.
First, the reference electrode data were analyzed.Figure 6 displays the reference electrode readings of Bolt No. 7 (E1, 455 Nm) over the 21 d span of the experiment with the Iin representing the electrical current through the ICCP system.Clearly, the reference-electrode reading became less negative as the electrical current faded, possibly because a low current implies fewer electrons in the electrical circuit.Consequently, fewer electrons congregated at the protected surface thereby providing weaker protection to the flange and the bolts compared to when the current was high and provided better defense against degradation.
Figure 7 shows the evolution of water temperature in the aquarium during the experimental timeframe of 3 weeks.Numbers 1 to 7 indicate the peak temperature on each day of the first week, which also help distinguish between the day-night temperature profile.Between the 21st and the 23rd of March 2021, an extremely hot weather spell raised the ambient temperature at Ashdod, Israel to 30 °C, 27 as reflected in the high water temperature in the aquarium at Points 5 to 7 in the plot of Fig. 7.
Temperature readings and reference-electrode measurements were concurrently plotted in Fig. 8. Interestingly, the aquarium's water temperature exhibited a strong correlation with the reference electrode measurements.When the water temperature ebbed during the night, the reference electrode readings became more negative than during the day when water temperatures were relatively high.

Discussion
As shown at Fig. 5, the magnitude of the torque significantly affected the corrosion rate of the bolts despite being shielded from corrosion by the ICCP.Past observations have revealed that the average corrosion rate of steel intensifies with increasing torque, regardless of the application or absence of the ICCP. 28Research results from this investigation are consistent with the previous findings which have demonstrated that material tension can enhance both the anodic and cathodic reactions of steel in a simulated marine atmosphere.When stress is elastic and steel operates well below its yield strength, its corrosion rate is hardly noticeable.We note that once the applied stress on the bolt reached at least 75% of the yield stress of steel, substantially enhance the electrochemical activity, promoting anodic dissolution while speeding-up the corrosion rate of steel. 29he second effect of the applied stress is stress concentration at the crack tip, which will magnify the plastic strain in the crack tip area and promote local anodic dissolution in that vicinity. 30Under an applied elastic-tensile stress, the cathodic reaction rate is considerably higher than in the absence of any load.Essentially, this indicates that application of mechanical stress can contribute more to the increase in the hydrogen evolution reaction than to the anodic steel dissolution reaction. 31he reference electrode reading (Eref), the electrical current (I in ), and the water temperature (T) measured in the aquarium are strongly related.Their correlations were observed over the 21 day-night temperature profiles.As the water temperature rose during daytime, the reference electrode readings became less negative, and the electrical current dropped.During nighttime, the opposite behavior was observed.Note that the water pump that boosted the water current remained active throughout the experiment and scattered the corrosion products while spurring the deposited layer.When the reference electrode reads less than −850 mV, the system becomes more susceptible to hydrogen embrittlement 31 and the water temperature strongly affects the ICCP performance.At higher temperatures, the anodic and cathodic reactions manifest faster, 32 along with the consumption of the electrons that protect the metal surface.Consequently, the absolute electron count at the metal surface shrinks (although limited by the impressed current) and the reference electrode reading becomes less negative.
In addition, the water temperature affects the mobility of the ions.At elevated temperatures, the water resistivity declines 23 and positive ions are easily and quickly passed to the metal surface, where they react with the free electrons inhabiting the protected configuration.In turn, the reaction rate on the metal surface intensifies and the electron count at the metal surface drops.Hence, the reference-electrode reading becomes less negative because of the fewer free electrons.
Finally, the water temperature governs the intensity of the electrochemical reactions.Rising aquarium temperatures foster water reduction via the HER given by Eq. 3. 33 Chemically, this is reflected by, the liberation of hydrogen gas molecules or hydrogen atoms at the metal surface.Under stress or applied torque, hydrogen atoms create the conditions for hydrogen embrittlement of the susceptible steel.Hydrogen embrittlement is a failure mechanism during which hydrogen diffuses to the metal surface. 34Hydrogen usually propagates along the grain boundaries, atomic vacancies, and dislocation tips, which eventually leads to the crack tip. 5 Under the combined effects of preload torque, water temperature, and activation, the ICCP system becomes susceptible to hydrogen embrittlement.When the ICCP configuration was activated, visible gas bubbles began forming on the devise surface.Visually, the gas bubbles accumulated across the flange and bolts configuration, as evidenced in Fig. 10.
The depth and characteristics of material cracks were observed through high-resolution SEM observations of the bolt samples.Small cracks were initiated at the surface of Bolt No. 7 (455 Nm).Inspecting the cross-section of this bolt, one can observe multiple brittle cracks which are apparently 2-4 times deeper than the   corrosion cracks (Fig. 11).The corrosion cracks extended only a few microns below the surface.Figure 12 displays the SEM images of Bolt No. 5 (341 Nm, 0.75 of Y. P.), which exhibit small cracks similar to Bolt No. 7 (Fig. 11).In contrast, Bolt No. 6 (114 Nm, 0.25 of Y. P.) showed no discernible cracks even under a magnification of 1,900.Apparently, the small torque magnitude applied to the fastened bolt was insufficient to crack it, either on the bolt surface or through the cross-section of the bolt (Fig. 13).
The depth of the crack in the protected bolts appears to be smaller, but an in-depth examination revealeds a morphological crack that points to hydrogen embrittlement. 35Clearly, while protecting the bolt from corrosion, more profound and yet detrimental hydrogen embrittlement cracks could emerge.

Conclusions
Undoubtedly, the mechanism of underwater corrosion is very complex.Concurrently, several marine related parameters, which vary from throughout the day t, constantly and simultaneously influence the underwater corrosion process.ICCP can improve the service conditions of subsea equipment.Meanwhile, the influence of subsea parameters such as temperature and the velocity of water currents on the ICCP performance should be considered.The main findings and their implications are summarized next: The torque magnitude of bolt tightening exerts a far-reaching effect on the corrosion rate of the bolts subject to the activated protection of the ICCP system.While the torque magnitude is raised from 114 Nm (25% of Y. P.) to 455 Nm (100% of Y. P.) the corrosion rate surged from 0.044 to 0.082 mm y −1 translating into an 87% faster corrosion rate.
Moreover, the water temperature greatly affects the combined effect of applied torque, water current, and water temperature that determine the nature of the underwater corrosion process.The diurnal profile of the water temperature affected the electric current and reference-electrode profiles during the 21 d of the experiment.Increasing temperature during daytime made the reference-electrode readings less negative while lowering the electrical current.Hence, lowering the cathodic protection performance.The ICCP system offered the lowest protection at the time of the peak temperature.
Meanwhile, the activated ICCP produced hydrogen molecules that were absorbed by the metal.Hydrogen embrittlement cracks were observed in the cross-sections of the bolts tightened at sufficient torque magnitudes (⩾0.75 times the yield strength of steel).If the material were to partially lose its ductility due to hydrogen embrittlement this also weakens its load bearing abilities in line with the failure of subsea critical bolts.

Figure 1 .
Figure1.Schematic configuration of the impressed current cathodic protection system displaying the water container together with the reference electrode, the cathode, and the anode.

Figure 2 .
Figure 2. Color illustration of the 8 bolts tightened under different torque magnitudes.

Figure 3 .
Figure 3. Picture displaying the powerful pump that simulates subsea water currents.

Figure 4 .
Figure 4. Flange-bolt configuration retrieved from the aquarium after 21 d.

Figure 9
compares the water temperature and the electrical circuit current (I in ) during the course of 21 d of the experiment.As the temperature fluctuated during the day and night, the electrical current measurements recorded by the ammeter displayed a similarly varying behavior, respectively.

Figure 5 .
Figure 5. Corrosion rates [mm y −1 ] of the bolts tightened under different torque magnitudes.

Figure 7 .
Figure 7. Water-temperature profile in the aquarium, numbers 1 to 7 indicate the peak temperature on each day of the first week and emphasize the day-night temperature profile.

Figure 8 .
Figure 8. Reference-electrode readings of Bolt No. 7 plot together with the water-temperature profile.

Figure 10 .
Figure 10.Bubbles suspected to be hydrogen accumulated on the surface of the bolts and flange during the experiment.

Figure 9 .
Figure 9. Electrical current readings of the ICCP superimposed on the water-temperature profile of Fig. 7.

Table I .
Torque magnitudes as percentages of yield point in the eight bolt-flange mechanical configurations.