ABSTRACT Corrosion of steel foundations in soil is a serious financial and technical problem for telecommunication companies. Underground corrosion is the primary cause of material degradation and structural failure at anchor shafts of guyed towers. Accordingly, accurate and practical methods to predict corrosion modes and corrosion rate are extremely beneficial for corrosion risk assessment and service life prediction. In this paper, field-proved guidelines for knowledge-based inspection, risk assessment, and risk mitigation of underground corrosion are highlighted which are specific to telecom structures. Effects of soil chemistry on corrosion of galvanized and carbon-steel anchor shafts are evaluated. A predictive model to estimate the corrosion rate and remaining service life of buried components are introduced. Four cases representative of different corrosion levels are presented and relevant mitigation process are given for each case. INTRODUCTION Guyed towers with anchor foundations is a common design for tall telecom towers that support antennas for broadcasting and telecommunication services such as radio, television, cellular networks, and satellite communications. Structural integrity of telecom towers is the key to ensure reliability of telecommunication and broadcasting services; nonetheless, many tower facilities are coming of age and corrosion related issues are turning into serious engineering and financial problems. These towers are suffering both from atmospheric corrosion at above-grade and underground corrosion at below-grade environments; however, the risk of structural failure is mostly associated with below-grade corrosion. An anchor foundation is schematically shown in Figure 1. The anchor shaft, usually made from carbon-or galvanized-steel, plays a critical role in structural integrity of guyed towers. This component is constantly under tensile force due to opposing forces at its ends, i.e., tension in guy wires and weight of the soil and anchor block. Accordingly, any form of material loss on anchor shafts would lead to decrease of cross-sectional area and formation of ‘stress risers’, which in turn would increase the risk of structural failure. This paper is a sequel to our previous NACE publication 1 , which was mostly focused on corrosion modes and cathodic protection system design. In the current paper, direct and indirect methods for corrosion risk assessment of anchor shafts are discussed, and sample case studies are presented. The direct approach includes visual examination and dimensional measurements, while the indirect method is based on soil characterization for corrosion rate determination and predictive modeling. The risk mitigation process is based on the identified risk factor.

ABSTRACT Predictive models: Corrosion rates can be obtained from predictive models considering various operating scenarios, i.e., before the onset of corrosion. Monitoring probes: Corrosion rates can be obtained from online monitoring probes as corrosion starts; due for example to operational changes and system upsets. Inspection (Inline inspection and non-intrusive probes): Corrosion rates can only be obtained only after corrosion has removed significant amounts of wall. INTRODUCTION Until 2000, the number of internal corrosion related incidences was more than 3 per 1,000 kilometers (KMs) of pipelines per year In 2017, the number of internal corrosion related incidences hovers around 0.5 per 1,000 KMs of pipelines per year In the future, the industry’s goal is to bring the number of incidences to “ZERO”. Predictive models: Corrosion rates can be obtained from predictive models considering various operating scenarios, i.e., before the onset of corrosion (KPI 12). Monitoring probes: Corrosion rates can be obtained from online monitoring probes as corrosion starts; due for example to operational changes and system upsets (KPI 24, 25, and 26). Inspection (Inline inspection and non-intrusive probes): Corrosion rates can only be obtained only after corrosion has removed significant amounts of wall (KPI 32 and 33).

ABSTRACT In this work, the peridynamic corrosion model is used for 3D simulation of pitting corrosion in stainless steel. Models for passivation and salt layer formation are employed to predict detailed characteristics of pit growth kinetic in stainless steels, such as lacy cover formation on top of the pit, and the diffusion-controlled regime at the pit bottom. The model is validated against an experimentally grown pit on 316L stainless steel in NaCl solution. Lacy covers in this model are formed autonomously during the simulation process. They are remarkably similar to the covers observed on top of the real pits. INTRODUCTION Corrosion in stainless steel (SS) is of significant interest due to the wide application of this alloy, with high mechanical strength and good corrosion resistance, in various industries. Pitting corrosion in SS takes the form of localized pits growing under perforated (lacy) covers. 1-6 Covered pits are difficult to spot, and they can maintain stable growth. 6 Since stable pits are potential sites for crack initiation, and can lead to catastrophic failure due to stress-corrosion cracking, 7 predictive models for pitting corrosion in stainless steel are of great importance. Such models could also be used to reveal the mechanisms of localized corrosion, and can help engineers and material scientists with finding new methods to stop or slow down damage progression under mechanical loading and environmental conditions. While many models exist for pitting corrosion, there exist a few that take into account the lacy cover on top of the pits in stainless steels. Most of these models include the lacy cover as a given, known boundary condition, and study pit growth under such covers, 8-11 . However, lacy covers influence pit growth and also are influenced by it, meaning that they should be obtained as part of the solving procedures, not enforced as pre-determined boundary conditions. There are only a couple of studies that model the actual formation of perforated pit covers. 12-14 Simulations in these studies are either in 2D, or axisymmetric, which are different from the general 3D conditions.

ABSTRACT Recently, several localized corrosion and cracking failures happened to API 5CT Grade C110 tubing in high-density formate completion fluid at high temperature, because of the leakage of CO 2 and the mechanical damage on the tubing surface. In this paper, the hydrochemical simulation software is used to calculate the needed content of pH buffer. The corrosion behavior of C110 tubing was investigated with autoclave, scanning electron microscopy (SEM) and X-ray diffraction (XRD). The results show that 1 wt% Na 2 CO 3 and 1 wt% NaHCO 3 can maintain the pH of formate completion fluid mixed with 3 MPa (435 psi) CO 2 up to 6.35, which means CO 2 corrosion is restrained effectively. The simulated experiments show that the general corrosion rate of C110 tubing decreases from 0.45 mm/year to 0.24 mm/year, and the C110 tubing shows little localized corrosion and cracking susceptibility in the formate completion fluid when buffered with 1 wt% Na 2 CO 3 and 1 wt% NaHCO 3 . The results of XRD and SEM indicate that the pH buffer is helpful to generate compact and protective FeCO 3 corrosion product film. INTRODUCTION The unique physic-chemical properties of high specific gravity, high temperature stability, high pH value and low corrosivity make formate solutions the ideal completion fluids. 1-3 However, recently there are several severe cracking and localized corrosion failures of API (1) 5CT Grade C110 tubing in high-density formate completion fluid at high temperature, because of large amount of acid gas (CO 2 ) influx and mechanical damage on the tubing surface. 4 CO 2 influx leads to the generation of carbonic acid and hence the decrease of pH value of formate brines solution. When pH is lower than 6.35, CO 2 corrosion will occur. 5-8 What’s worse, the lower pH, the greater the tendency for generating and decomposing of formic acid. 9-12 The main ways to reduce the risk of tubing corrosion and cracking in the completion fluid are improving the completion fluid’s ability of maintaining the pH values, enhancing the sealing performance of coupling to prevent the CO 2 influx and reducing the mechanical damage to the tubing. In this paper, the pH buffer of Na 2 CO 3 and NaHCO 3 is used to stabilize the pH value of the formate brine completion fluid. The effect of pH buffer to the contaminated formate brine completion fluid is investigated by simulation calculation and autoclave test. Also the corrosion products of corrosion tests are studied by X-ray diffraction (XRD) and scanning electron microscope (SEM).

ABSTRACT Studies in microbiologically influenced corrosion (MIC) have reported on the effects of pre-corrosion surface deposits on the localized pitting which occurs on metals. But due to the complexity and heterogeneity of these deposits, which include biofilms, it is necessary to investigate how the components of these deposits and the conditions therein influence the formation of pits on the metal surfaces. To gain a better understanding of the occurrence and growth of pits under these deposits, it is imperative to consider their interactions with the metal surface at the atomistic level. In this work, molecular modelling is used to study these interactions, with the focus being on parameterizing the role of HS - in microbiologically influenced pitting. The bond length of HS - is used as a predictive parameter in the molecular model to study the MIC interface. It is observed that changes in the HS - bond length denote HS - reactivity and the subsequent production of sulfides which are the main by-products of MIC. This study also shows how changes in temperature impact HS - reactivity and thus MIC activity. INTRODUCTION Since microbiologically influenced corrosion (MIC) was identified as one of the main threats to asset integrity management in the oil and gas industry, numerous studies have been done to understand the reactions that occur at the corrosion interface. 1-4 While these studies have contributed to the knowledge base with respect to MIC, there are still knowledge gaps and contradictions between studies. 56 This is largely due to the complexity of the deposits which form on the metal surface. These surface deposits include inorganic compounds from the environment and biofilms which accumulate on metals before pitting occurs. 7-9 In addition to its microbial communities, surface deposits on metals are composed of a heterogenous mixture of large organic molecules such as polysaccharides, proteins, lipids and nucleic acids, smaller inorganic species produced as intermediates and end-products of corrosion, metal deposits, and other metabolic products. 7,10 This results in chemical reactions which can impact the thermodynamics and kinetics of MIC development, thus making the process more difficult to understand.

ABSTRACT Most of the corrosion prediction models used for design of oil and gas lines carrying high pCO 2 are valid up to 1 ~ 2 MPa of pCO 2 and are very conservative at higher pCO 2 because they do not account for the effect of high pCO 2 on the water chemistry and the corrosion mechanism. The present work was focused on developing a predictive tool for near-critical and supercritical CO 2 corrosion of mild steel. It incorporates changes in the water chemistry module due to update solubility and dissociation equations, changes in the electrochemical module due to the presence of a thick and porous corrosion product layer, and consideration of an adsorption mechanism for H 2 CO 3 at the steel surface. The comparison between experimental results and model predictions showed a good agreement under various pressure and temperature ranges. INTRODUCTION Internal corrosion of carbon steel pipeline in high partial pressure CO 2 (pCO 2 ) environments have gained more interest recently relating to carbon capture and storage (CCS), enhanced oil recovery (EOR), and deep water oil and gas production applications.1-5 Although there are extensive research data available on high pCO 2 corrosion, minimal information has been reported in the literature that could aid in establishing a corrosion model for carbon steel in such conditions.6-8 Furthermore, the published literature on supercritical CO 2 primarily addresses topics related to sequestration and enhanced oil recovery (EOR) that usually involve “dry” gases where water is only present at the ppm level. The impact of aqueous CO 2 corrosion on carbon steel has been studied extensively at pressures relevant for oil and gas transport (up to 2 MPa CO 2 ). Therefore, most of the predictive models used for design of oil and gas lines carrying high pCO 2 are valid up to 1 ~ 2 MPa of pCO 2 and are very conservative at higher pCO 2 . 9-12 It has been reported that aqueous corrosion mechanisms in high pCO 2 are similar to those in low pCO 2 conditions including characteristics of corrosion product layer and the response of corrosion rate with temperature. 13 This indicates that the existing CO 2 corrosion model could be used to predict the corrosion rate at high pCO 2 conditions with some modifications in chemical, electrochemical or transport processes. For example, since CO 2 changes from gaseous to liquid or supercritical with increasing pressure, it will lead to different interaction with water, i.e. CO 2 solubility in water will not follow Henry’s law in liquid or supercritical CO 2 conditions, which results in changing water chemistry. 14

ABSTRACT Top of Line Corrosion (TLC) occurs in a multiphase flow when water vapor condenses at the top and the sides of the pipeline, leading to a severe corrosion attack. This study investigated the probabilistic risk of TLC for wet sour gas subsea pipeline using flow modeling and corrosion predications. The flow assurance hydraulic study showed that most of water drops out over the first few kilometers as the gas is cooled and becomes much less through the rest of the offshore part until they reach onshore area where the gas temperature drops further due to Joule-Thomson effect. It was anticipated that corrosion activities will be higher at the high condensation locations. The corrosion prediction modeling revealed high corrosion severity driven by Top of Line Corrosion (TLC). In order to maintain the system integrity the internal coating supplemented by V-jet batch inhibitor injection has been selected to protect against TLC. This study has realized the challenge to apply the batch treatment as it requires process interruption to meet scraper speed limitations. Therefore, the industry path forward should consider the development of novel TLC treatments to be applied with no impact on operations. INTRODUCTION The hydrocarbon exploration in the ocean and deep sea started as early as 1850s, when the first drilling was carried out in California. Other early oil exploration activities were later recorded in Pakistan (1886), Peru (1869), India (1890) and Dutch East Indies (1893). 1 In 1930s, the development of the Gulf of Mexico as an offshore area started with oil first being produced in 1938. The production from the North Sea brought more technical challenges in to the offshore industry. The North Sea was first explored as a potential offshore area in the early 1960s. 2 Since then the pace of oil exploration and production in shallow water has gradually increased to deep water with the exploration phase starting in 1975 while production began twenty years later. 2

ABSTRACT ABSTRACT Environmental exposure history, measured using corrosion sensors, is used to drive a computational model and predict the cumulative galvanic corrosion damage. Knowing the environmental exposure of an aircraft over the course of its lifetime is critical to making meaningful predictions of accumulated corrosion damage. Dynamic environmental monitoring using sensor technology provides a significant advantage compared to using historical atmospheric data records. Estimates of time-of-wetness (TOW) are made through statistical manipulation of the sensor data so that discrete environmental characterizations (WET, SEMI-WET and DRY) of the total exposure history can be derived. Computational models require information on electrolyte conductivity, film thickness, and polarization properties to make predictions of galvanic stress and metal loss. By combining TOW data with electrolyte properties representing each of the discrete environmental exposure conditions, and using corrosion rates from computational models, the cumulative corrosion damage for a defined operational period can be predicted. A demonstration of this new methodology (corrosion Service Life Model) is presented using actual environmental exposure data collected onboard a naval vessel. The effects of crevice and pitting corrosion can also be included in this predictive model. The approach is applicable not only to aircraft but to a wide range of structures subject to atmospheric corrosion. INTRODUCTION Aircraft are exposed to a wide range of environmental conditions over their operational life and this influences the type and severity of corrosion damage they experience. For example, sea based aircraft experience a much more aggressive, salt-laden, environment than land based aircraft. In order to make meaningful predictions of cumulative corrosion damage it is critical to know the environmental exposure of an aircraft over the course of its lifetime. Material degradation mechanisms such as galvanic, pitting, crevice, and intergranular corrosion limit the operational lifetime of aircraft and result in unsustainable long term maintenance costs. Recent studies of naval aircraft suggest that galvanic corrosion is the primary corrosion mechanism, in nearly 80% of the cases, when corrosion damage is observed on the aircraft. The annual cost of corrosion for Navy and Marine Corps aviation is estimated at $3.4 B, accounting for 27.9% of the total maintenance budget, with annual non-availability due to corrosion at 228,471 days. Corrosion maintenance is often very labor intensive so more effective scheduling can reduce costs; there is an economical need to move from a Find and FixÅapproach to a more efficient and cost-effective Predict and PreventÅ approach.

ABSTRACT Predicting scale formation; Identifying the upper driving force limit for inhibitors and blends; Developing inhibitor models for minimum effective dosage; Developing models for preventing failure due to inhibitor solubility; and Modeling inhibitor synergy and competitive inhibition. INTRODUCTION scale predictions are based upon simple indices. 1,2,3,4 indices are calculated from total analytical values and do not account for the ion associations and common ion effects prevalent in higher dissolved solids waters. calculations estimate carbonate from uncorrected “M” alkalinity titrations and do not correct for non-carbonate alkalinity. ion activity estimations use methods appropriate for low ionic strength solutions, but that lack accuracy and applicability in high total dissolved solids (TDS) systems.

ABSTRACT Accurate quantification of corrosivity of crude oil fractions has been a major challenge for refiners, stemming from inadequate understanding of complex influences of naphthenic acids and type/ morphology, molecular weight of acid species, sulphur content and speciation, temperature, and fluid flow on corrosion. A proprietary model has been developed encapsulating data from Honeywell’s Crude Corrosivity Phase-I Joint Industry Program (JIP), conducted between 2006 and 2011. This research program, sponsored by Indian Oil Corporation Limited (IOCL) and nineteen other global refining and engineering companies, led to the development of the first ever quantitative engineering database and decision-support model to predict corrosion for common materials of construction employed in high temperature refinery crude fractionation. Data were generated in simulated refinery environments for relevant temperature ranges, varying naphthenic acid content, and active Sulphur levels for different process oils. This paper provides details from the evaluation of the proprietary prediction model and validation studies conducted in IOCL’s refinery crude units. Case studies comparing the model predictions with the refinery inspection / measured corrosion rate data as a function of key operating parameters are provided. Guidelines on the utility of the prediction model and suggestions for improvements are also provided. INTRODUCTION The processing of such opportunity crudes offers refinery operators significant economic benefits while posing serious challenges in terms of potential for corrosion damage of unit equipment as well as product quality. Some of the barriers to opportunity crude processing stem from naphthenic acid corrosion, fouling, sulphidic attack, crude compatibility, long term accessibility, crude quality variation, increased hydrogen demand, coke make, reduced product yield of distillates, increased energy demand and increased CO 2 emissions. These issues can be addressed in a long-term perspective by design modification, metallurgy upgrading, and adoption of relatively new resid upgrading processes. New refineries can be designed accordingly depending on the premised crude slate. However, older refineries that were designed to process light sweet crude (API > 30°, S < 0.5Wt. with low TAN < 0.5mgKOH/gm) face the challenge of remaining operationally viable when faced with the daunting task of processing high acid and reactive sulphur crudes [2,3]. Against the backdrop of availability of the opportunity crudes, at least over the medium term and in regard to the concomitant economic benefits of refining such crudes or crude blends, it becomes imperative, especially for the older refineries, to look at the challenges and draw short and long-term plans for processing opportunity crudes. This paper evaluates the corrosion behaviour of naphthenic acid bearing crudes and the quantification of corrosion rate of susceptible locations using the crude prediction model. The evaluation and validation has been achieved utilizing several real case studies at IOCL refineries, wherein the performance of the prediction model has been compared against inspection-based corrosion measurements [1-3].

ABSTRACT Environmentally assisted cracking of high strength nickel based alloys under was investigated in 3.5wt% NaCI under cathodic polarization. Rising displacement tests were performed on C22HS and MP98T were performed at a slow K-rate of 0.005Nmm -3/2 /s. The results suggest that there is no significant environmental effect for MP98T, however C22HS exhibited a decrease in the initiation toughness. Tests performed under K control under a range of conditions on 718 and K-500. In general, the crack growth rate (CGR) measured under constant K were lower than those obtained under rising displacement tests under similar conditions. The CGR in both 718 decreased with decreasing potential. The CGR was also observed to depend strongly on the loading mode. There was a strong effect of increasing K, with increasing K tests resulting in substantially higher CGR’s than at constant under similar conditions. The above results suggest that the rate limiting step for crack propagation may be the generation of hydrogen at the crack tip. A crack tip strain rate based model was applied to rationalize the data though more work is needed to establish the various parameters associated with the model. INTRODUCTION High strength nickel based alloys like Monel K-500, Inconel 718, Inconel 625+/725 have been known to be susceptible to hydrogen embrittlement/1-3/. Slow strain rate tests revealed susceptibility to intergranular cracking in K-500 in seawater over a range of applied cathodic potentials from -850mV to - 1000mV SCE/4/. Field failures of K-500 were attributed to hydrogen embrittlement in seawater under cathodic charging/5/. It was suggested that in these cases that high hardness associated with cold work from rolling threads on fasteners may have been the primary cause of enhanced susceptibility to hydrogen embrittlement/5/. The presence of carbon films along the grain boundaries was also attributed to increased susceptibility, though the exact mechanism that is operative was not clear/6/. Constant load and double-cantilever-beam (DCB) testing performed on UNS N07718 and UNS N07725 demonstrated that under cathodic polarization of -1 V to -1.1 V SCE (in synthetic seawater) these materials are resistant to hydrogen embrittlement/3/. However, the observation of failures in the field suggest that there may be other factors responsible for cracking which were not simulated in the published work. It is possible that in constant load tests the absence of a pre-crack coupled with dynamic straining may be a reason for not observing cracking. Slow rising displacement tests (K-rates of ~0.3 to 0.5MPavm/h) have been employed recently to evaluate susceptibility of nickel based alloys to hydrogen embrittlement in seawater under cathodic polarization/711/. The results indicate that materials such as K-500, and Inconel 718 exhibit susceptibility under these conditions, with K th values significantly lower than measured in dead loaded/rising step loaded tests/7,9/. The slow rising displacement tests have been valuable in trying to discern susceptibility as measured via K th . However, it is interesting to note that in a number of systems evaluated using the slow rising displacement method/7-12/, the measured steady plateau CGR even at low K-rates is on the order of 10” 6 to 10 -4 mm/s/8-12/. More recent work that has demonstrated that the CGR’s of Inconel 718 are sensitive to K-rates/7/ and when measurements were performed at K-rates significantly lower than 0.5MPaVm/6/. Similar data has been observed for K-500 when comparing the CGR data under rising displacement versus constant load conditions/7/. In this context, it is important to note the measured CGR’s at K-rates of 0.5MPaVm/h may not be meaningful for developing damage tolerant designs (CGR of 10 -6 mm/s would result in a crack extension of ~1inch/year). Under these conditions it is clear that hydrogen embrittlement is present, it is important to note that any predictive model needs to address the observed sensitivity to K-rate. Recent efforts to model the measured CGR have essentially focused on the relating the measured CGR to the effective diffusivity of H, and the concentration of hydrogen, without accounting for the observed strain rate dependencies in any manner/10-12/.

ABSTRACT The main purpose of this joint industrial R&D project was to perform detailed field measurements on selected pipeline segments in close vicinity to a High Voltage Direct Current (HVDC) transmission line during planned staged fault tests on the power line. Induced voltages on these pipelines were recorded during the HVDC staged fault process, which was conducted by the utility company. The measurement results from these tests were used for further validation or modification of an existing industrial guideline which focuses on the influence of HVDC power lines on metallic pipelines. Verification of future modelling results was another expected outcome of this study. A pre-field laboratory test was conducted to ensure accurate calibration and functionality of equipment prior to the actual staged fault field testing. To check the functionality of the digital triggering on the data logging unit and choose an appropriate triggering level and methodology, a typical touch voltage waveform based on previous simulation forecasts was used. The test setup during the HVDC staged fault testing consists of an arrangement of ground rods, a differential high voltage probe and the data logging unit used in order to capture the transient fault waveforms on the pipelines. INTRODUCTION We conducted detailed induced voltage measurements on selected pipeline segments affected by a High Voltage Direct Current (HVDC) transmission line during planned staged faults on this power line. A detailed HVDC interference simulation study was also completed on these selected pipelines to evaluate induced touch voltages during staged fault condition. These pipelines were chosen based on the staged fault test locations and their relative geometry with HVDC power line. A comparison between measured and simulated results for further validation or modification of an existing industrial HVDC guideline, and verification of future modelling results were the primary concerns in this project. HVDC transmission lines can couple in three different ways to parallel conductors: through mutual capacitance, mutual inductance, and through direct conduction.

ABSTRACT Atmospheric corrosion represents an annual multi-billion dollar cost burden for the aerospace and defense sectors. For many aircraft, particularly those operating in marine environments, up to ninety percent of corrosion is due to galvanic interactions at dissimilar metal couples. As new materials are introduced with the acquisition of more advanced aircraft, galvanic corrosion is likely to remain a concern. The ability to model galvanic corrosion accurately holds the promise of being able to both predict the performance of new material combinations to guide material selection and predict corrosion damage for maintenance planning. Such models often utilize data collected under immersion test conditions that are not representative of the thin-film electrolytes that are relevant to atmospheric corrosion and may diminish model accuracy and utility. In this work, an atmospheric cell is presented that allows for measurements of corrosion kinetics using thin-film electrolytes. It is observed that the limiting oxygen reduction current density on various alloys is increased several orders of magnitude over immersion results. A segmented, galvanic sensor is presented that enables the experimental quantification of spatial distributions of galvanic current under thin film conditions that is compared to model predictions for verification of the suitability of immersion and thin-film electrolyte polarization data inputs. INTRODUCTION Corrosion represents a significant risk and cost burden for the aerospace and defense industry. For example, atmospheric corrosion of U.S. Navy and Marine Corps aircraft, alone, results in costs of approximately $2.6 billion each year. The primary atmospheric degradation mode for aircraft is galvanic corrosion at dissimilar metal couples (e.g. between non-aluminum metallic fasteners and the aluminum structure) [1]. This problem will become even more pronounced as new material combinations are introduced in advanced aircraft. Current maintenance practices are schedule-based leading to either unnecessary over-inspection or worse, unplanned and costly downtime when excessive corrosion is observed. There exists a need for advanced tools and methods to both improve the prediction of material performance at galvanic couples and corrosion damage under atmospheric conditions. With better predictive capabilities, maintenance and sustainment of aircraft can be optimized to decrease costs and increase aircraft availability. Computational models offer a solution to help address this need; however, current predictive models of material performance at galvanic couples are based upon data collected in non-representative conditions. This may diminish the accuracy and utility of the models to guide corrosion resistant material selection during acquisition and corrosion prediction throughout the aircraft’s lifetime.

ABSTRACT Industry guidance such as European Standard EN-15280 (1) and applied research by Pipeline Research Council International (PRCI) (2) highlight the interaction cathodic direct currents can have on pipeline corrosion influenced by alternating currents. Often, excessive levels of cathodic protection (or cathodic interference) can exacerbate AC-influenced corrosion. Conversely, effective management of cathodic protection in conjunction with monitoring the AC and DC effects on a pipeline can be a valued component of an optimized AC corrosion control strategy. This paper presents case histories where throttling down the cathodic protection was evaluated to determine the impact on reducing the AC corrosion threat. The investigations include the use of fast-response electrical resistance (ER) corrosion rate probe monitoring technology. INTRODUCTION This paper presents four case histories relating to monitoring and mitigating AC-influenced corrosion: • Case History Nos. 1, 2 and 3 convey evaluations to determine how controlling cathodic protection levels can help mitigate AC-influenced corrosion. • Case History No. 4 illustrates the value of remote monitoring to detect upset conditions and act upon them as appropriate. The example is associated with pipeline corrosion control impacted by the deluge rainfall that hit the Gulf Coast region of Texas and Louisiana during Hurricane Harvey in August-September 2017. There are countless combinations of DC effects, AC effects, and soil characteristics that can interact and impact AC-influenced corrosion. This paper does not delve into the detailed electrochemistry or kinetics associated with this interaction, other than to simply state that AC-influenced corrosion is governed heavily by Ohm’s Law at the pipe surface-soil interface. Influencing the voltage, current and or resistance of the circuit determines the corrosion significance. For details on the theoretical foundation for the interactions between DC and AC the reader is referred to References 1 and 2 and the various other literature cited in these references. CASE HISTORY NO. 1 – 100+ MILES, 6 RECTIFIERS This case history covering over 100 miles (161 km) of fusion bonded epoxy (FBE) coated pipe was the subject of a NACE International 2017 conference paper. The challenging pipe route traverses the massive Atchafalaya Swamp and extensive agricultural tracts in the Gulf Coast region of southern Louisiana and Texas. Elevated AC interference levels on the pipeline from 45 paralleling and crossing high voltage AC (HVAC) power lines ranging to 230 kilovolts (kV) prompted a comprehensive AC investigation.

ABSTRACT Industry research over the last decade has demonstrated the potential for numerical simulation, such as Monte Carlo, to project future corrosion of fixed equipment in refineries and chemical plants. This research has shown the potential for numerical simulation to provide metallurgists and inspection engineers advanced tools in managing fixed equipment in refineries and chemical plants. This paper expands on this work by first reviewing the basis for utilizing Monte Carlo simulation to predict future corrosion in fixed equipment. This paper then explores implementing Monte Carlo simulation based on the quality of inspection data by proposing methods for high-quality, statistically-based inspection data as well as historic, non-statistically-based data. This paper proposes how to incorporate the simulation results into a risk based inspection plan aligned with the philosophy of industry risk based inspection practices. INTRODUCTION An ever increasing number of inspections are being performed at refineries and chemical plants in order to manage mechanical integrity due to the wide adoption of risk based inspection (RBI). These programs are typically organized to be in compliance with API 581, which provides simplified guidance for conducting inspection to obtain a quantitative result that can be compared in a seriatim against all other inspection results so that end users can prioritize remediation appropriately. 1 This simplification can mask the uncertainty inherent in risk management through inspection. Occasionally the results of risk based inspection identify a component or piece of equipment with either a high consequence of failure (from a safety or financial perspective) or a risk where appropriate follow up inspection is high cost or results in adverse impact to unit operations. These specific instances provide an opportunity for a statistically-based model to clarify the risk. Materials and inspection engineers can use this information to facilitate better life-cycle management discussions for developing a business case that supports appropriate future inspection or equipment replacement.

ABSTRACT This study was conducted to determine optimal coupon size and shape of the coupons with minimal edge effect in terms of cathodic current distribution. Specifically, experiments and corrosion modeling were conducted to develop an approach to evaluate optimal cathodic protection (CP) coupon size and shape for monitoring CP in field. Experiments were conducted with 10, 50, and 100 cm 2 coupons in circular, triangular, and cylindrical shapes. The experiments included linear polarization resistance, potentiodynamic cathodic polarization, and potentiostatic polarization measurements. The experimental cells used in the experiments were modeled for the current distribution at the coupons surfaces. The pH of the sand directly next to the coupons surface was measured, which was found to be more basic than the bulk sand pH, indicating that cathodic currents resulted in generation of alkalinity which was absorbed by the sand. The experimental data were input in a pipeline CP model that included a 10 cm 2 holiday and a coupon with varying sizes and shapes. The coupon sizes and shapes included a factorial combination of 10, 50, and 100 cm 2 surface areas and circular and cylindrical shapes. The simulation data indicated that edge effects are minimal for a 10 cm 2 cylindrical coupon and 100 cm 2 circular coupon under various cathodic polarization conditions. INTRODUCTION Cathodic protection (CP) in combination with coatings is broadly used to prevent or mitigate external corrosion of buried pipelines, tank bottoms, and other applications. In the United States (U.S.), the current NACE International Standard SP0169-2013 Section 6 mainly emphasizes three CP criteria: (i) -850 mV CSE on-potential, (ii) -850 mV CSE off-potential (i.e., polarized potential), and (iii) 100 mV polarization. 1 CP coupons are used extensively in the industry to assess protection levels. They vary widely in shape and size. Sizes range from 1 to 100 cm 2 and the shapes are (i) square or round flat, (ii) long slender rod, (iii) small rod with cylinder at one end, (iv) stubby cylinder and others. NACE Standard SP0104-2014 states that ”The proper coupon design must be selected,” but provides only limited information on the significance of the design parameters. 2

ABTRACT A comprehensive corrosion prediction model for multiphase oil and gas production and transmission systems was enhanced by incorporating a Top-of-Line corrosion capabilities. Key issues in development of the enhancements are discussed, including evaluation of an improved flow model implemented to facilitate accurate flow predictions. The model was evaluated with parameters commonly encountered in oil and gas production environments as well as transmission systems, and effects of various parameters on corrosion rates are described. Corrosion rates obtained from the enhanced model are compared with actual field and lab testing data as a basis to quantify accuracy and efficacy of the model. INTRODUCTION Although Top of the Line Corrosion (TLC) was observed as early as 1960, the potential damage to the oil and gas transportation lines was well recognized only in the past one or two decades 1 . Compared to typical aqueous corrosion, the phenomenon of TLC is understood to a much less extent, especially in H 2 S dominant environment. Research efforts undertaken in various research institutes and companies have shed some lights on the mechanisms and key factors governing TLC in CO 2 environment. Several theories and methodologies have been proposed and developed to predict CO 2 Top of the Line Corrosion 1-10 . Top of the Line corrosion typically occurs in carbon steel pipelines/flowlines in which wet gas or stratified multiphase flow are transported. A significant temperature difference between the fluids inside the line and the ambient environment can lead to TLC initiation. Field experiences have suggested that continuous corrosion inhibitor application is usually ineffective in TLC scenario. This is because corrosion inhibitors tend to stay in the liquid phase and are unable to reach the top portion of the pipe in stratified flow. Moreover, for wet gas pipelines, it is usually very challenging to transport corrosion inhibitors in long distance and ensure complete coverage of internal pipe surfaces. As condensation occurs down the line, corrosion takes place and can exacerbate by the presence of acid gases, such as CO 2 and acetic acid, which dissolves in the condensed water to generate exceptionally low pH.

ABSTRACT Evaluation of the performance of cathodic protection (CP) systems used on subsea structures once in service can be costly due the need for specialized personnel and ROV monitoring equipment. Often areas of a subsea structure can go unassessed in terms of CP coverage when the ability to assess the entire structure is not feasible. Improperly protected areas can lead to premature corrosion or structural integrity issues, or costly maintenance and retrofits. The use of computer software can aid in identifying improperly protected areas. Boundary element analysis (BEA) of a CP system, validated with past inspection data, can produce a computer model that will allow for predictive assessments of the present and future states of the CP system and subsea structure. Identification of under/over protection ‘hot spots’ by using a BEA model can guide the focus of the next inspection for a more thorough assessment of critical locations. In this study, the CP system of an offshore complex consisting of Al-Zn-In galvanic anodes was modeled using BEA. The offshore complex consists of concrete and steel jackets, risers and pipework. Time-stepping, anode degradation and coating breakdown are incorporated to estimate system performance with time since installation. The results of the CP system simulation were validated against existing inspection data and used to optimize the established CP inspection program. INTRODUCTION An offshore complex consisting of concrete structures and steel jackets, risers and pipework operates in the North Sea. The subsea portion of the main platform is comprised of a concrete gravity oil storage tank, concrete tower nearly centered atop the concrete tank and a steel jacket and conductor frame. The concrete tank connects to the lower portion of a steel jacket which rises vertically to create the second connection point to the topside platform structure along with the concrete tower. The first of two ancillary steel jackets reside near the main platform and connects via subsea pipeline to the second ancillary steel jacket approximately 2 kilometers away.

ABSTRACT Due to the electrical conductivity of pyrrhotite it was hypothesized that its presence in the corrosion product layer on a steel surface could lead to localized corrosion. Mild steel specimens (API 5L X65) were pretreated to form a pyrrhotite layer on the surface using high temperature sulfidation in oil. The pretreated specimens were then exposed to a range of aqueous CO 2 and H 2 S corrosion environments at 30 and 60°C. X-ray diffraction data showed that the pyrrhotite layer changed during exposure; in an aqueous CO 2 solution it underwent dissolution while in a mixed CO 2 /H 2 S solution it partially transformed to troilite, with some mackinawite formation. This led to initiation of localized corrosion in both cases. Propagation of the localized attack was enhanced due to a galvanic coupling between the pyrrhotite layer and the steel surface. The intensity of the observed localized corrosion varied with solution conductivity (NaCl concentration); a more conductive solution resulted in higher localized corrosion rates consistent with the galvanic nature of the attack propagation. INTRODUCTION In H 2 S containing environments encountered in the oil and gas industry, localized corrosion is a potential cause leading to facility failure. There can be a high rate of metal loss in a very limited area, which may be covered by a corrosion product layer. This makes H 2 S localized corrosion more difficult to predict and detect prior to failure by using the conventional corrosion inspection and monitoring methods.1,2 Considering the often random spatial distribution of localized attack and the limited number of monitoring probes that can be installed in any given facility, the chances of detecting localized corrosion this way are slim at best. Internal line inspection techniques which could theoretically detect localized attack are complicated, expensive and therefore are used infrequently. Thus, a better understanding of localized corrosion mechanisms would be essential for the development of predictive models and implementation of corrosion mitigation strategies. There are complicating factors associated with the investigation of H 2 S corrosion mechanisms. This includes the recently found electrochemical mechanisms involving direct reduction of H 2 S at the metal surface 3-5 and the role of different iron sulfides 6-12 that can form on the metal surface in the corrosion process. In H 2 S solutions, the corrosion product layer can be composed of various iron sulfides with distinct physicochemical and electrical properties 13-18 . The electrical conductivity of various iron sulfides is one of the key parameters. For example, pyrrhotite (Fe 1-x S), troilite (FeS) and pyrite (FeS 2 ) all occur as stable corrosion products and have similar electrical conductivities 13-15 while for the more unstable mackinawite (FeS) there are far fewer values reported for its conductivity 19 . Mackinawite has anisotropic electrical properties, being conductive in the direction of oriented layers in its crystal structure and much less conductive in the perpendicular direction 19 . The existence of conductive phases on a steel surface significantly impacts the electrochemically driven corrosion process 3,20,21 . The conductive corrosion product layer may intensify the electrochemical reaction rate through providing a larger cathodic surface area, locally or uniformly across the corroding steel surface.

ABSTRACT The environments of chemical plant facilities are complex and multiple corrosion factors are frequently intermingled. When corrosion takes place, these factors interact and may camouflage the underlying causes of the corrosion damage, making it difficult to separate and elucidate them. However, in continuous processing facilities, the time element is a useful additional factor that often can enable different corrosion phenomena to be tracked and their cause(s) identified. In order to take advantage of this situation, real time monitoring techniques are applied to clarify the contributory causes of corrosion attack that is sustained from time to time by the plant during normal operation. In this report of practical on-line monitoring, the use and benefits of the EN instrumentation, in combination with process chemistry analyses, were employed to monitor operating conditions in a full-scale Coke Oven Gas (COG) desulfurization process in which the corrosion rate changed with time. 1 The underlying causes of corrosion and the effectiveness of possible countermeasures were evaluated using corrosion simulation software. As a result of progressive improvements, the severe corrosion attack that had been sustained by the facility was reduced to less than 0.05 mm/y after formulation and application of the countermeasures. INTRODUCTION Chemical plants provide society with a wide variety of products through reactions that change the raw materials into useful substances. In many cases, as a variety of feedstocks may be employed, the combination of products generated by the chemical reactions may differ significantly, as shown in Figure 1. When corrosion takes place, these factors interact and the combination tends to hide the underlying causes of corrosion damage. However, as shown in Figure 2, in continuous processing facilities, the time element is a useful additional factor that often can enable different corrosion phenomena to be tracked and their causes attributed to specific process environments to be identified. In order to take advantage of the time element, real time monitoring techniques are applied to clarify the contributory causes of corrosion attack that may occur from time to time during normal operation. In this report of practical on-line monitoring, the use and benefits of the electrochemical noise (EN) instrumentation, in combination with process chemistry analyses, were employed to monitor operating conditions in a full-scale Coke Oven Gas (COG) desulfurization process in which the corrosion rate varied with time.