Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Metallurgical investigation of premature failures for 316L austenitic stainless steel pipes Al Muaisub, Mohammed 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_september_almuaisub_mohammed.pdf [ 4.21MB ]
Metadata
JSON: 24-1.0366239.json
JSON-LD: 24-1.0366239-ld.json
RDF/XML (Pretty): 24-1.0366239-rdf.xml
RDF/JSON: 24-1.0366239-rdf.json
Turtle: 24-1.0366239-turtle.txt
N-Triples: 24-1.0366239-rdf-ntriples.txt
Original Record: 24-1.0366239-source.json
Full Text
24-1.0366239-fulltext.txt
Citation
24-1.0366239.ris

Full Text

METALLURGICAL INVESTIGATION OF PREMATURE FAILURES FOR 316L AUSTENITIC STAINLESS STEEL PIPES by Mohammed Al Muaisub   B.Sc., King Fahd University of Petroleum and Minerals, 2007  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE  in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Materials Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)      May 2018  © Mohammed Al Muaisub, 2018 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the thesis entitled:   Metallurgical investigation of premature failures for 316L austenitic stainless steel pipes  submitted by Mohammed Al Muaisub  in partial fulfillment of the requirements for the degree of Master of Applied Science in Materials Engineering  Examining Committee: Edouard Asselin, Materials Engineering Supervisor  Tom Troczynski, Materials Engineering Supervisory Committee Member  Daan Maijer, Materials Engineering Supervisory Committee Member   iii  ABSTRACT This thesis analyzed and investigated the premature failures of pipes made from type 316L austenitic stainless steels. Multiple leaks were observed in scattered locations of a piping network of around 10 km after only 4 months in service transferring ammonium sulfate solution. The initial investigation indicated that the piping network was constructed 3 years earlier. After the construction, the stainless steel pipes were hydrotested to ensure the joints integrity. However, the piping network was not properly drained and dried after the hydrotest which resulted in water stagnation for the complete idle period between construction and commissioning. Therefore, an electrochemical, chemical, mechanical and metallurgical testing and analyses were conducted to determine the damage mechanism which consequently caused these failures. I have conducted electrochemical tests on a 316L stainless steel electrode in chloridized ammonium sulfate solution to determine its corrosivity. The electrochemical tests showed that the corrosion rates of 316L SS in ammonium sulfate solution is very low. This conclusion was supported by other laboratory studies at higher temperature and by the industrial corrosion tables published online. Also, two spools from the piping network that experienced the failures were analyzed using stereoscope, optical microscope, scanning electron microscopy/energy dispersive spectrometry, X-ray fluorescence and carbon/sulfur analyzer, tensile testing and microhardness testing. The results of these tests indicated that the pipes were leaking at the 6 O’clock position near the weld and heat affected zone areas. The morphology of the attack illustrated a narrow opening with large sub-surface cavity and tunneling initiated from the internal surface of the pipes. The weld joints displayed weld defects in terms of root concavity and lack of penetration. The iv  metallurgical investigation strongly suggests that the pipes failed due to Microbiologically Influenced Corrosion (MIC). During the idle period of 3 years, the stagnant untreated water in the closed system was an appropriate environment for bacterial growth leading to severe damage at the welding joints and the base metal.    v  LAY SUMMARY This thesis discussed and investigated an interesting type of failure that many industrial plants, as well as commercial and residential buildings, might experience in their water system network. Pipes, or even equipment, made of stainless steel, might fail in an accelerated manner due to bacteria. The scientific term for this type of failure is Microbiologically Influenced Corrosion (MIC). These bacteria do not “eat” the metal per se, but they are directly or indirectly involved in chemical and electrochemical reactions that increase the rate of the attack on the steel components. It is reported that MIC is a rather common type of corrosion damage that might represent up to 20% of the total corrosion failures in stainless steel systems.       vi  PREFACE  This thesis is original, independent work by the author, Mohammed Al Muaisub. The electrochemical tests and analysis that were presented in Chapter 5 is based on my work in the Corrosion Lab in UBC. The metallurgical, chemical and mechanical tests in Chapter 7 were conducted in the laboratory of the sponsoring company. The following was extracted from the analytical and research work which is presented in the thesis.  Conference:  Mohammed Al Muaisub, Akram Alfantazi, "Premature Failure of 316L Stainless Steel Pipes Due to MIC", 2nd NACE European Area Conference, 27-29 May 2018, Genoa, Italy (Poster Presentation - Submitted and Accepted)     vii  TABLE OF CONTENTS ABSTRACT ..................................................................................................................... iii LAY SUMMARY .............................................................................................................. v PREFACE .......................................................................................................................vi TABLE OF CONTENTS ................................................................................................. vii LIST OF TABLES ............................................................................................................xi LIST OF FIGURES ......................................................................................................... xii LIST OF SYMBOLS, ABBREVIATIONS, AND NOMENCLATURE ............................... xvi ACKNOWLEDGEMENTS ........................................................................................... xviii 1 INTRODUCTION ...................................................................................................... 1 2 BACKGROUND ........................................................................................................ 3 3 LITERATURE REVIEW .......................................................................................... 10 3.1 Austenitic Stainless Steels ............................................................................... 10 3.2 Pitting Corrosion in Type 316L SS ................................................................... 13 3.2.1 Critical Factors in Pitting Corrosion ........................................................... 14 3.2.1.1 Surface Environment ................................................................................. 15 3.2.1.2 Potential ..................................................................................................... 16 3.2.1.3 Alloy Composition ...................................................................................... 16 3.2.1.4 Temperature .............................................................................................. 17 3.2.1.5 Surface Condition ...................................................................................... 18 viii  3.2.2 Corrosion in Ammonium Sulfate Solution .................................................. 18 3.2.3 Pitting Corrosion in Potable and Sea Water .............................................. 20 3.3 Microbiologically Influenced Corrosion (MIC) ................................................... 22 3.3.1 Susceptible Materials to MIC ..................................................................... 23 3.3.2 Mechanisms of MIC Bacteria Types .......................................................... 24 3.3.2.1 Sulfate Reducing Bacteria (SRB) .............................................................. 25 3.3.2.2 Iron Oxidizing Bacteria (IOB) ..................................................................... 27 3.3.3 Morphology of the MIC Attack ................................................................... 27 3.3.4 Locations of MIC Failures .......................................................................... 28 3.3.5 Effect of Water Quality on MIC .................................................................. 31 4 OBJECTIVES ......................................................................................................... 33 5 ELECTROCHEMICAL TESTING ON 316L SS IN AMMONIUM SULFATE SOLUTION .................................................................................................................... 34 5.1 Specimen Preparation ...................................................................................... 34 5.2 Electrolyte Composition ................................................................................... 35 5.3 Corrosion Cell Setup ........................................................................................ 35 5.4 Electrochemical Techniques ............................................................................ 36 5.4.1 Open Circuit Potential (OCP) ..................................................................... 36 5.4.2 Potentiodynamic Polarization Test ............................................................ 36 5.5 Results and Discussion of the Electrochemical Tests on 316L SS Electrode .. 37 ix  5.5.1 Effect of Increasing Chloride Concentration .............................................. 37 5.5.2 Effect of Increasing Ammonium Sulfate Concentration ............................. 42 5.5.3 Effect of Increasing the Scan Rate of the Anodic Polarization Test ........... 43 6 TESTING PROCEDURE FOR THE FAILED 316L SS PIPES ................................ 45 6.1 Visual Examination ........................................................................................... 45 6.2 Chemical Testing ............................................................................................. 46 6.2.1 Chemical Analysis of the Process Sample ................................................ 46 6.2.2 X-Ray Fluorescence and Carbon/Sulfur Analyses ..................................... 46 6.3 Mechanical Testing .......................................................................................... 48 6.3.1 Tensile Tests ............................................................................................. 48 6.3.2 Hardness Tests ......................................................................................... 50 6.4 Metallurgical Testing ........................................................................................ 51 6.4.1 Samples Preparation for Metallography Tests ........................................... 51 6.4.2 Stereoscope and Optical Microscope Analyses ......................................... 52 6.4.3 Scanning Electron Microscopy and Energy-Dispersive X-Ray Spectroscopy Analyses ................................................................................................................. 53 7 RESULTS AND DISCUSSION FOR THE FAILED 316L SS PIPES ....................... 55 7.1 Visual Examination ........................................................................................... 55 7.2 Chemical Testing ............................................................................................. 64 7.2.1 Chemical Analysis of the Process Samples ............................................... 64 x  7.2.2 X-Ray Fluorescence and Carbon/Sulfur Analyses ..................................... 64 7.3 Mechanical Testing .......................................................................................... 65 7.3.1 Tensile Tests ............................................................................................. 65 7.3.2 Hardness Tests ......................................................................................... 66 7.4 Metallurgical Testing ........................................................................................ 68 7.4.1 Stereoscope and Optical Microscope Analyses ......................................... 68 7.4.2 Scanning Electron Microscopy and Energy-Dispersive X-Ray Spectroscopy Analyses ................................................................................................................. 72 7.5 Effect of Welding on MIC Initiation ................................................................... 88 7.6 Effect of Hydrostatic testing procedure on MIC Initiation .................................. 89 7.7 Effect of the Idle period between construction and commissioning on MIC propagation ................................................................................................................ 90 8 CONCLUSIONS AND RECOMMENDATIONS ....................................................... 92 8.1 Summary of the Findings from the Conducted Laboratory Tests ..................... 92 8.2 Summary of Key Observations from the Investigation ..................................... 95 8.3 Recommendations to Avoid the Reoccurrence of the Failure .......................... 96 8.4 Suggested Future Work ................................................................................... 97 BIBLIOGRAPHY ........................................................................................................... 99 APPENDIX .................................................................................................................. 114    xi  LIST OF TABLES Table 12.1: Process and Operation Parameters ............................................................. 5 Table 22.2: Findings of the Conducted NDT Methods by the Plant Inspectors ............... 6 Table 33.1: Chemical Composition of Type 304L and 316L SS (ASTM A312, 2016). .. 12 Table 43.2: Mechanical Properties of 304L and 316L Austenitic Stainless Steels. (ASTM A312/A312M, 2016) ...................................................................................................... 12 Table 53.3: Pitting Resistance Equivalent Number for type 304L and 316L Stainless Steels ............................................................................................................................ 17 Table 65.1: Testing Environment and Conditions .......................................................... 35 Table 76.1: The Dimensions of the Specimens Used in the Tensile Tests .................... 49 Table 87.1: pH and chloride concentrations in 3 process samples ............................... 64 Table 97.2: XRF and carbon-sulfur analysis for the two failed pipes ............................. 65 Table 107.3: Tensile tests results illustrate that both samples meet the mechanical properties requirements for the type 316L SS. .............................................................. 66 Table 117.4: Microhardness values for the failed sample including the weld area, HAZ and parent metal. .......................................................................................................... 67 Table 127.5: Qualitative EDS analysis for the S1 spot. Note the high carbon content suggesting the presence of bacteria.............................................................................. 80 Table 137.6: Qualitative EDS analysis for the S2 spot. ................................................. 81 Table 147.7: Qualitative EDS analysis for deposit Area. ............................................... 83 Table 157.8: Qualitative EDS analysis for the area specified in Figure 7.31. ................ 86 Table 167.9: Qualitative EDS analysis for the area specified in the following Fig 7.32. 86  xii  LIST OF FIGURES Figure 12.1: Photographs of the leaking pipes. Note that the pinholes located at or near the weld joints. Note the white flaky deposits from the leaked ammonium sulfate solution. ........................................................................................................................... 7 Figure 22.2: Photographs of the leaking pipes. Note that the pinholes here are located at the parent metal. ......................................................................................................... 8 Figure 32.3: Photographs of two X-Ray films generated using RT inspection method at the weld joints pinholes’ locations. Note the rounded indications alongside the weld area. ................................................................................................................................ 9 Figure 55.1: Open Circuit Potential of 316L SS in 2 – 4 g/l of chloride in 100 g/l ammonium sulfate solution ............................................................................................ 38 Figure 65.2: Open Circuit Potential of 316L SS in 6 – 8 g/l of chloride in 100 g/l ammonium sulfate solution ............................................................................................ 38 Figure 75.3: Open Circuit Potential of 316L SS in 400 g/l ammonium sulfate solution .. 39 Figure 85.4: The effect of increasing chloride content from 2 - 4 g/l in 100 g/l ammonium sulfate on the 316L working electrode ........................................................................... 40 Figure 95.5: The effect of increasing chloride content from 6 - 8 g/l in 100 g/l ammonium sulfate on the 316L working electrode ........................................................................... 41 Figure 105.6: The effect of presence of the chloride ions on the corrosion behaviour of the 316L working electrode in 400 g/l ammonium sulfate solution ................................ 41 Figure 115.7: Open Circuit Potential of 316L SS in chloridised ammonium sulfate solution at 1 g/l of NaCl ................................................................................................. 42 xiii  Figure 125.8: The effect of increasing the ammonium sulfate concentration from 100 to 400 g/l on the 316L working electrode........................................................................... 43 Figure 135.9: Anodic polarization curves show the effect of increasing the scan rate in the experiment on the measured potentials and current densities. ............................... 44 Figure 146.1: The Tensile Test Sample. Note the actual dimensions in Table 6.1. ....... 49 Figure 157.1: Photograph of the failed spool in as received condition. Note that the pinhole located at the weldment area. ........................................................................... 56 Figure 167.2: Photograph of the other failed spool in as-received condition. Note that the pinhole here is located at the parent metal. ............................................................. 56 Figure 177.3: A Close-up view of the failed spool showing the pinhole at the fusion line. ...................................................................................................................................... 58 Figure 187.4: A Close-up view of the failed spool showing the pinhole at the parent metal. ............................................................................................................................ 58 Figure 197.5: Close-up view of the internal surface of the failed spool. ........................ 59 Figure 207.6: Photograph shows the pinhole and the subsurface tunneling after sectioning the spool circumferentially. Note the close-up views of the pinhole mouth at the ID. ............................................................................................................................ 60 Figure 217.7: A Close-up view of the sectioned spool near the pinhole. ....................... 61 Figure 227.8: Photograph shows the internal surface of the failed spool. ..................... 62 Figure 237.9: Close-up view shows the multiple pits initiated at the incomplete root penetration area of the weld as indicated in Figure 5.8. ................................................ 62 Figure 247.10: Tensile test results show the tensile and yield strength of the two samples. ........................................................................................................................ 66 xiv  Figure 257.11: Photomacrograph showing the exact location of the indentations for the microhardness testing. .................................................................................................. 68 Figure 267.12: Photograph shows a prepared and hot mounted sample of the localized damage at the weld area. .............................................................................................. 69 Figure 277.13: Photomicrograph shows the localized attack in as polished-condition. . 70 Figure 287.14: Photomicrograph shows the localized attack in as etched-condition. .... 70 Figure 297.15: Photomicrograph of the 316L SS spool sample away from the pinhole in as-polished condition. .................................................................................................... 71 Figure 307.16: Photomicrograph of the 316L SS sample away from the pinhole in as-etched condition. ........................................................................................................... 72 Figure 317.17: SEM image at low magnification showing the cross-section photograph of the subsurface cavity and tunneling. 20x Magnification. ........................................... 73 Figure 327.18: SEM image showing the cross-section photograph of the subsurface cavity. Note the preferential attack mechanism. 70x Magnification ............................... 74 Figure 337.19: SEM image showing the cross-section photograph of the subsurface cavity. Note the preferential attack mechanism. 100x Magnification ............................. 74 Figure 347.20: SEM photomicrographs showing the cross-section of the subsurface cavity. Note the preferential attack mechanism. 250x Magnification ............................. 75 Figure 357.21: Photographs show the morphology of MIC attack on the fusion line of SS weldment. Note the dendritic attack as pointed by (D) compared to nondendrite area on (A). The Image on the right shows the MIC attack of the fusion line and the weld area. (Jack, 2002) .................................................................................................................. 76 xv  Figure 367.22: SEM image shows the pinhole in as received condition. Magnification 25x. ............................................................................................................................... 77 Figure 377.23: A Magnified SEM image of the same location of Figure 7.9. Note the dimension of the pits openings. Magnification 100x. ..................................................... 77 Figure 387.24: SEM image shows the same previous pinhole at higher magnification (200x). Note the clear preferential attack mechanism at the edge of the pit. ................. 78 Figure 397.25: SEM images of the cross-section of the pinhole in as-received condition. Note the irregular shape suggesting bacteria presence. Mag. 1200x and 2000x. ......... 79 Figure 407.26: A form of bacterial cells in a colony at a magnification of 2700x. (Dexter, 2003) ............................................................................................................................. 80 Figure 417.27: EDS analysis for the S1 spot in Figure 7.12. Note the high carbon peak. ...................................................................................................................................... 80 Figure 427.28: EDS analysis for the S2 spot in Figure 7.12. Notice the high iron peak and the low carbon peak. .............................................................................................. 81 Figure 437.29: SEM Image shows deposits on the internal surface of the pipe. ........... 82 Figure 447.30: EDS analysis of the doposits as shown in Figure 7.29. ......................... 83 Figure 457.31: EDS analysis (Above) for the pinhole in as-received condition (Below). 85 Figure 467.32: EDS analysis (Above) for the pinhole in another location in as-received condition (Below). .......................................................................................................... 87        xvi  LIST OF SYMBOLS, ABBREVIATIONS, AND NOMENCLATURE T Temperature icorr Corrosion Current Density Ecorr Corrosion Potential Ep Pitting Potential Er Repassivation Potential MAWP Maximum Allowable Working Pressure RT Radiographic Testing UT Ultrasonic Testing NDT Non-Destructive Testing MIC Microbiologically Influenced Corrosion UDC Under-Deposit Corrosion NPS Nominal Pipe Size ID Inner Diameter OD Outer Diameter (NH4)2SO4 Ammonium Sulfate Solution NaCl Sodium Chloride NACE National Association of Corrosion Engineers API American Petroleum Institute ASTM American Society For Testing Materials XRF X-Ray Fluorescence SEM Scanning Electron Microscopy EDS Energy Dispersive X-Ray Spectrometer   NH3 Ammonia H2SO4 Sulfuric Acid Ø Diameter xvii  SS Stainless Steel ppm  Parts Per Million PREN Pitting Resistance Equivalent Number CPT Critical Pitting Temperature Cl SCC Chloride Induced Stress Corrosion Cracking SRB Sulfate Reducing Bacteria H2S Hydrogen Sulfide IOB Iron Oxidizing Bacteria CMIC Chemical Microbiologically Influenced Corrosion EMIC Electrical Microbiologically Influenced Corrosion HAZ Heat Affected Zone WPS Welding Procedure Specification OCP Open Circuit Potential PDP Potentiodynamic Polarization ACS American Chemical Society Ag/AgCl The Silver/Silver Chloride Reference Electrode LVDT Linear Variable Differential Transformer PT Dye Penetrant Test OM Optical Microscope ISO International Organization for Standardization ASM American Society for Metals HRB Rockwell Hardness Scale B HV Vickers Hardness TDS Total Dissolved Solids TSS Total Suspended Solids    xviii  ACKNOWLEDGEMENTS  I offer my continuing gratefulness to my supervisor Dr. Alfantazi for his guidance and support that helped me in my graduate study. His feedback and patience helped me to complete successfully my thesis and thus my Master Degree in Materials Engineering.     I would like also to thank Dr. Asselin, Dr. Maijer, Dr. Poole and Dr. Troczynski who served as the examining committee members for my thesis. Their constructive feedback tremendously helped me to improve my thesis. My thanks goes to the faculty, staff and my fellow students in materials engineering department who I learned a lot from them all.    Special thanks to my colleagues Matlub, Ibrahim and Mohammed who assisted and motivated me on a continuous basis.   Also, thanks to my former manager in SABIC Mr. Ali Al-Hazemi who provided the support and encouragement to continue my higher education. Finally, thanks to Saudi Basic Industries Corporation (SABIC) who supported me financially to continue my graduate studies abroad in the University of British Columbia (UBC).1  1 INTRODUCTION In petrochemical industries, one of the major roles of metallurgical and corrosion engineers is to investigate and analyze different types of failures that were experienced in piping and pipe fittings, tanks, vessels, and rotating equipment in the plants. One of the main objectives of such investigations is to clearly define and identify the possible damage mechanisms that contributed to the failure. Also, an equally important objective of any such investigation is to propose mitigation methods or remedial actions in order to avoid the reoccurrence of the failure in the future.   Conducting a failure analysis is a complex process that requires broad knowledge of different technical disciplines and the use of different techniques. For example, it is essential to know the difference between an indicator, a cause, a damage mechanism and a consequence.  Although the previously mentioned terminologies might overlap, it is critical to be able to recognize them during the failure analysis process in order to determine the suitable course of action that will help prevent the repetition of the failure in the future. (Scutti, 2002)  Recently, a petrochemical plant experienced a major incident in one of the main pipes that carries ammonium sulfate solution. There were multiple leaks due to localized attack in different locations along the pipe. This localized attack was in the form of numerous pits in the heat-affected zone and base metal. The pit morphologies and locations, as well as the history of the implemented hydrotesting procedure and water flushing activities, 2  suggested that the probable damage mechanism was Microbiologically Influenced Corrosion (MIC). Since the total length of the pipes is around 10 kilometers and its material of construction is type 316L austenitic stainless steel (SS), the financial impact of this failure was several million dollars. In addition, the piping network was just 4 months in service at the time of failure. All of these facts indicated that this was a high-value failure analysis project. Moreover, because of the widespread leaks, such an investigation became a high priority due to the adverse effects on maintaining the continuous operation of the plant.   So, the question here is why did a newly commissioned 316L stainless steel pipe system leak in multiple locations after only 4 months in service? In addition, how did it happen? Therefore, the aim of this study is to investigate and analyze the damage mechanism of the newly installed 316L stainless steel pipe system. Moreover, a thorough review of the literature was presented on the localized damage mechanisms of austenitic stainless steels with a specialized focus on Microbiologically Influenced Corrosion (MIC). In addition, the literature review chapter highlights the adverse effect of the weldment and prolonged water stagnation on the corrosion behaviour. Also, a detailed collection of data and discussion were presented in the background chapter about the pre-commissioning and commissioning procedures, the process and operating parameters and the conducted Non-Destructive Tests (NDT) conducted by the plant inspectors.  3  2 BACKGROUND  It is crucial, as a part of the metallurgical failure investigation process, to collect all the information and documents that are related to the failure in question. This important step is required since the failure may not be a direct consequence of a single cause. Understanding this point will broaden the horizon of the failure analyst trying to unearth all possible causes that might contribute to the failure at hand. For example, when a tube failed due to stress corrosion cracking, two conditions should be present in order to have this damage mechanism. The first condition is the presence of an environment where there should be chemical or electrochemical reactions take place on the surface of the material i.e. corrosion mechanism. The second condition is residual or applied tensile stresses. Without these conditions, stress corrosion cracking is not possible.   Therefore, in this chapter, all related information about the failed pipes was presented. This information and data includes the material of construction of the failed pipes, the location of the failure in the plant site, Non-Destructive Testing (NDT) results conducted on the site, construction and commissioning related data, and finally, process and operating parameters.  The pipes material of construction, which are intended to transfer ammonium sulfate solution, are made from type 316L austenitic stainless steel. The nominal pipe size “NPS” is 3 inch schedule 10, i.e. the pipe outer diameter is 3.5 inch and the pipe thickness is 0.12 inch (3.05 mm). The pipes were in service for only 4 months. However, the piping 4  was first constructed 3 years earlier. The pipes were joined together by welding using ER316L as the filler material which has similar chemical composition to the base metal. The industrial standard used for the pipe installation required that upon completion of construction, and to ensure the integrity of the pipes, hydrostatic testing should be carried out as part of the pre-commissioning procedure. The main purpose of the hydrostatic testing, or in short hydrotesting, is to ensure that the pipe system can be operated at the Maximum Allowable Working Pressure (MAWP). Another purpose is to ensure that there are no major defects in both the material and the welded joints. After hydrotesting, the pipes should be drained, dried, purged with nitrogen and kept in this preserved condition for the entire extended idle period.   Plant documents showed that hydrotesting was conducted. However, the water quality and proper draining and drying were not controlled and monitored. In addition, during the 3 year idle period, purging with an inert gas like nitrogen was not performed.    The quality of the water used for the hydrotesting should follow the specified guideline. Seawater and raw/well waters should not be used. Even potable water could only be used if was treated to ensure that chloride, oxygen and microorganisms were minimal. During the review of the background information on this failure, no document was found pertaining to the type or the quality of the water used.   5  Table 2.1 shows the process media and the operating parameters for the solution transferred through the pipes after commissioning the plant.   Table 12.1: Process and Operation Parameters  Description Value 1 Process concentration 10% ± 1.0 wt. ammonium sulfate and 0.2% Urea 2 Ammonium sulfate pH 3.5 as a set value with performance limit between 3-4 3 Ammonium sulfate density 1050 Kg/m3 normal/set value with performance limit of 1000-1100 Kg/m3 4 Temperature 45oC – 50oC 5 Pressure 3 – 4 Kg/cm2g (42.7 – 56.9 psi)  In this plant, ammonium sulfate is produced from a chemical reaction between ammonia and sulfuric acid, as shown in the reaction below: 2𝑁𝐻3  +  𝐻2𝑆𝑂4  =  (𝑁𝐻4)2𝑆𝑂4 When the concentration of the ammonium sulfate solution reached 10% wt., the solution is pumped to an ammonium sulfate crystallization unit through a number of pipes. In that unit, ammonium sulfate crystals are produced. The failed pipes were part of the piping circuits that carried the produced solution to the ammonium crystallization unit. As part of evaluating the criticality of the leakage, the plant inspectors conducted a series of NDT methods on the leaking pipes in situ as shown in Table 2.2. 6   Table 22.2: Findings of the Conducted NDT Methods by the Plant Inspectors  NDT Method Findings 1 External Visual Inspection Multiple pinholes leaks were observed in the weldments and in the parent pipe. All the leaks were located at the bottom of the pipe, 6 O’clock position. No jetting or dripping of fluid was noticed from the leak locations and only minor wetness & solid deposit (white flaky deposit) was observed around the leak location. Also, all the weld joints on the line were found to be with minor rust. However, no other abnormality noticed on the line. See Figures 2.1 and 2.2.  2 Ultrasonic Testing (UT) Ultrasonic Testing (UT) to measure the pipe thickness was carried out and all the readings were found to be acceptable; for 3” Pipe Ø: UT readings are between 2.8 mm to 3.20 mm against 3.05 mm Nominal Thickness. 3 Radiographic Testing (RT) Rounded indications (of size 3~6 mm dia) and elongated indications (up to 10 mm long) were observed in many weld joints along the circumference. No other abnormality noticed on the line. See Figure 2.3.   7   Figure 12.1: Photographs of the leaking pipes. Note that the pinholes located at or near the weld joints. Note the white flaky deposits from the leaked ammonium sulfate solution. Weld Joint Weld Joint Pinhole Pinhole The white buildup deposits are from the leaked ammonium sulfate The white buildup deposits are from the leaked ammonium sulfate 8    Figure 22.2: Photographs of the leaking pipes. Note that the pinholes here are located at the parent metal. Pinhole Pinhole The white buildup deposits are from the leaked ammonium sulfate 9    Figure 32.3: Photographs of two X-Ray films generated using RT inspection method at the weld joints pinholes’ locations. Note the rounded indications alongside the weld area.     The light-colored region is the circumferential weld joining the pipes Indications of weld defects Indications of weld defects 10  3 LITERATURE REVIEW Based on the background of this failure, the 316L stainless steel pipes experienced localized corrosion attack. Also, the abnormal idle time between construction and commissioning is anticipated to play a significant role in the failure mechanisms. Moreover, since the pinholes were located at or near the welded area, this indicates that welding procedure/parameters could also contribute to this failure.   In this chapter, the published literature was reviewed in the following areas: localized attack of 316L SS in the form of pitting corrosion, the corrosion behaviour of 316L SS in ammonium sulfate solution, the corrosion of weldments in stainless steel, the MIC attack of 316L SS and its common relation with hydrostatic testing in terms of the water quality, the procedure, and the aftermath of the extended idle period.   3.1   Austenitic Stainless Steels Since petrochemical industrial equipment and piping deal with different chemicals and service environments with different operating parameters i.e. temperature, pressure, pH, flow rate, etc., it became necessary to select suitable materials of construction that can maintain such conditions. Austenitic stainless steels can withstand processes with temperatures varying between cryogenic levels up to 600 oC. In addition, they are suitable to be used in severe corrosive environments (Davis, 2006).  11  The family of austenitic stainless steels comprises many alloys with different chemical compositions. Generally, they are classified as a family of iron based alloys that have more than 50% of Iron, between 16% to 26% of chromium and less than 35% of nickel (Demeri, 2013). The two most common alloys in this category are 304L SS and 316L SS, where the “L” stands for low carbon content. The advantage of using the L grades is to maintain the corrosion resistance of the material by reducing its susceptibility to sensitization (AWS D10.4, 1986). The sensitization occurs due to the precipitation of the chromium carbide along the austenite grain boundaries which results in depleting the adjacent boundaries from chromium that will increase the susceptibility to corrosion (Kutz, 2002).   The 304L and 316L SS differ in the chromium and nickel contents, which increase the corrosion resistance and the hardenability of the materials, respectively. Yet, the major difference between the two alloys is the addition of molybdenum in the composition for the resistance to halogen acids. Molybdenum increases the pitting resistance of the material from oxidizing chlorides. Table 3.1 shows the chemical composition of the two alloys. (Kelly, 2002)    12  Table 33.1: Chemical Composition of Type 304L and 316L SS (ASTM A312/A312M, 2016). Grade Weight, %, maximum, unless otherwise indicated C Mn Si P S Cr Ni Mo TP 304L 0.035 2.00 1.00 0.045 0.030 18.0-20.0 8.0-13.0 --- TP 316L 0.035 2.00 1.00 0.045 0.030 16.0-18.0 10.0-14.0 2.00-3.00  As for the mechanical properties, austenitic stainless steels normally have superior toughness, higher rates of work hardening and low yield strength with high ductility when compared against ferritic or martensitic stainless steels (Cramer, 2005). Table 3.2 shows the mechanical properties of the 304L and 316L austenitic stainless steels. Notice that the minimum requirements for the tensile and yield strengths are affected by the carbon content.   Table 43.2: Mechanical Properties of 304L and 316L Austenitic Stainless Steels. (ASTM A312/A312M, 2016) Grade UNS Designation. Tensile  strength, min ksi [MPa] Yield  strength, min ksi [MPa] Elongation in 2 in. or 50 mm (or 4D), min, % Longitudinal Transverse TP304 S30400 75 [515] 30 [205] 35 25 TP304L S30403 70 [485] 25 [170] 35 25 TP316 S31600 75 [515] 30 [205] 35 25 TP316L S31603 70 [485] 25 [170] 35 25  As mentioned earlier, one of the main purposes for the use of austenitic stainless steel is its ability to resist a corrosive attack in certain environments. Olsson (2003) stated that the presence of the passive layers on stainless steel surfaces gives such material its 13  superior corrosion resistance characteristic (Olsson, 2003).  It is worth noting that these passive layers on the stainless steel surfaces are influenced by the alloying elements in the chemical composition (Olsson, 2003). Once the bare surface is exposed to the environment, the chemical reaction between them results in a thin and resistant film made of chromium oxide that drives the corrosion rate to very low level (Olefjord, 1980).   Generally, various damage mechanisms could take place on piping and equipment made from type 316L austenitic stainless steels. In rare cases, stainless steel could experience uniform corrosion. However, the most common forms of corrosion are as follows: pitting, crevice corrosion, intergranular corrosion, galvanic corrosion, microbiologically influenced corrosion, high-temperature corrosion, stress corrosion cracking, and erosion (Dillon, 2015).  In the subsequent sections, a greater focus was placed on one of the major forms of localized corrosion, which is pitting corrosion. Then, in a separate section, a detailed literature review was presented on Microbiologically Influenced Corrosion (MIC).  3.2   Pitting Corrosion in Type 316L SS In the beginning, it is critical to correctly characterize the localized corrosion and pitting as to whether they are corrosion forms or distinctive damage mechanisms. In order to do so, localized corrosion and pitting should be defined first. Localized corrosion is generally a loss of metal that occurs in isolated areas on surfaces that are generally uncorroded. 14  On the other hand, pitting could be classified as an extreme form of localized corrosion and can be defined as loss of metal with a common morphology of a V shape, i.e. penetrates deeply with a narrow surface opening. Therefore, pitting should be considered as a corrosion form rather than separate damage mechanism. This is mainly due to the various damage mechanisms that exhibit pitting as the morphology of the attack (Dillon, 2015).    As mentioned previously, stainless steels are selected to be used in many static and rotating equipment due to a thin oxide layer called the passive film in nanometer-scale which is formed on the surface and significantly lower the corrosion rate. However, the passive layers can breakdown in localized locations. When this happens, it results in an accelerated corrosion of the underlying metal which then causes pitting. The criticality of these pits could be understood by their consequences. For example, main structural components could fail due to major localized thickness loss that adversely affect the minimum required thickness to maintain the working pressure. Also, pits could lead to major components failure by acting as initiation sites for different cracking mechanisms. (Frankel, 2003)  3.2.1 Critical Factors in Pitting Corrosion  There are various factors influencing pitting corrosion. Such factors include surface environment, surface condition, metal composition, potential and temperature. Critical surface environment factors include ion concentrations, inhibitor concentrations, and pH. 15  Other fundamental aspects of the pitting corrosion are the stochastic nature of the processes and the different stages of the localized attack. Therefore, in this section, a thorough review was presented on these critical factors. (Frankel, 2003)  3.2.1.1 Surface Environment Generally, pitting corrosion is a result of the attack of aggressive anion species on the oxide passive layer (Frankel, 2003). Such attack is exhibited as localized damage due to the passive layer breakdown (Frankel, 2003). Moreover, it was well established in the early stages of studying the pitting corrosion that the severity of this type of corrosion is strongly correlated with the bulk chloride concentrations (Leckie, 1966). This is due to the fact that chloride is an anion of a strong acid where a number of metal cations show significant solubility in chloride solutions (Galvele, 1981). Generally, type 316L stainless steel is suitable for use at ambient or near ambient temperature for low concentrations of chloride, i.e. in the range of 200 – 1000 ppm (Grubb, 2005).  The corrosion inhibitor needs to be introduced first prior to the initiation of any pits. Controlling the pitting using corrosion inhibitor once the pit started might not solve the problem. Pitting is thought to be autocatalytic in nature which means that once the pit initiated and started growing, the localized environment is changed in a way that promotes further growth of the pit (Frankel, 2003).  16  3.2.1.2 Potential All metals and alloys have characteristic potentials. Generally, for a typical austenitic stainless steel potentiodynamic polarization test result., the scan starts from lower potentials to higher potentials where a pit will initiate when the potential reaches the pitting potential (EP). The pit will carry on to propagate with high current density until the potential lowers to reach the repassivation potential (ER) where the passive film will reproduce, thus significantly lowering the current density. (ASTM International Standard G61 - 86, 2014)  3.2.1.3 Alloy Composition As highlighted by Szklarska-Smialowska (2004), the composition of the alloy and the microstructure could be controlled to decrease the susceptibility of the material for pitting. Moreover, the physical and chemical imperfections of the materials which are formed during the production of most engineering alloys could lead to the pits development. Various elements like molybdenum, chromium, nitrogen and nickel were stated to have a significant effect on improving the pitting resistance of stainless steels. (Szklarska-Smialowska, 2004) It is worth noting that a simplified yet effective equation was introduced to determine the pitting resistance of various materials based on the alloying elements composition. The Pitting Resistance Equivalent Number (PREN) could be calculated based on the following equation: (Cleland, 1996) 𝑃𝑅𝐸𝑁 = 𝐶𝑟 (𝑤𝑡. %) + 3.3𝑀𝑜 (𝑤𝑡. %) + 16𝑁 (𝑤𝑡. %) 17  Lorenz first introduced the basic form of this equation in 1969 as follows: (Lorenz, 1969) 𝑃𝑅𝐸𝑁 = 𝐶𝑟 + 3.3𝑀𝑜 However, it was noticed that nitrogen plays a significant role in the pitting resistance when added as an alloying element. Therefore, Truman in 1987 fine-tuned the PREN equation to be the currently most used format as highlighted earlier with the addition of 16N (Truman, 1987). Table 3.3 illustrates the PREN values for the materials listed in Table 3.1.  Table 53.3: Pitting Resistance Equivalent Number for type 304L and 316L Stainless Steels Grade UNS Designation Cr Mo PREN TP 304L S30403 18.0 – 20.0 --- 18 – 20 TP 316L S31603 16.0 – 18.0 2.0 – 3.0 22.6 – 27.9  3.2.1.4 Temperature Temperature role in determining the pitting resistance is vital. Based on many studies, a new correlation with pitting potential was introduced which was labeled as Critical Pitting Temperature (CPT) where it derived from the stipulated experimental procedure for each metal and alloy. It was found that in artificial seawater, for example, the pitting potential decreases when the solution temperature increases. Therefore, materials that demonstrate higher CPT tend to have higher resistance to pitting corrosion. (Ovarfort, 1989; Arnvig, 1996)  18  3.2.1.5 Surface Condition Another mostly unconsidered critical factor for pitting corrosion is the material surface condition where materials with rougher surface exhibit higher susceptibility and rate for pitting corrosion. It was reported that a stainless steel sample of type 302 that was finished to 120-grit had a pitting potential of about 150 mV lower than a sample that was finished to 1200-grit in chloride containing environment. (Laycock, 1997)  Moreover, Sedriks (1996) demonstrated in his publication “Corrosion of Stainless Steels” that multiple applied treatments like heating, grinding and abrasive blasting could detrimentally affect the pitting resistance. As for the heat treatment, if applied improperly, it could result in a chromium-depleted region that introduced a preferential site for pitting due to the lower corrosion resistance. Furthermore, another side effect of improper heat treatment is the heat tint oxide that also becomes a preferred site for pitting to initiate. (Sedriks, 1996)  3.2.2 Corrosion in Ammonium Sulfate Solution A limited number of papers and studies were found discussing and investigating the electrochemical and corrosion behaviours of 316L stainless steels in ammonium sulfate solutions at low temperature environment. One of the earliest published studies on the corrosion behaviour in ammonium sulfate solution was conducted by Sugibayashi in 1954. Although the article was published in the Japanese language, the summary was translated into English. It was found that the corrosion rate of 18-8 steel, i.e. 304 SS, 19  decreases at temperatures below 80 oC and significantly increases at temperatures higher than 90 oC. They also reported that when the 18-8 material was modified by the addition of molybdenum, the corrosion rate decreased. In particular, with the addition of 2.65 % Mo, the corrosion rate decreased to less than 5% of that measured for 18-8 based on a 7-day immersion test, i.e. corrosion loss in weight of 0.11 mg/cm2 vs. 2.6 mg/cm2 (Sugibayashi, 1954).  Another early study on the corrosion behaviour of ammonium sulfate solution at evaporative conditions was conducted by Lebedev (1974). The laboratory tests were conducted in a harsh environment to simulate the worst-case scenario in the industrial setting where the ammonium sulfate solution was at a boiling temperature and at pH of 5. The Kh18N10T and Kh17N13M2T materials, equivalent to 304 and 316 stainless steels were exposed to such conditions for up to 4.5 months. It was found that the corrosion rate of Kh18N10T is 0.11 mm/year whereas for Kh17N13M2T, the corrosion rate was 0.07 mm/year. Such findings coincide with the previously mentioned study that shows the better corrosion resistance of 316 SS comparing to 304 SS. (Lebedev, 1974)       On another note, one of the closely related studies was conducted by Ghahreman (2012), on 316L and 317L SS alloys in chloride containing ammonium sulfate (3.56 M) solution at 100 oC. It was found that in naturally aerated condition, 316L is adequate material in neutral and slightly acidic environments i.e. pH between 5.4 and 7.4 with the presence of chloride at 0.12 M. (Ghahreman, 2012)  20  The same goes while reviewing the industrial guidelines, Outokumpu, a large corrosion resistant alloy manufacturer, has published online its own corrosion tables for materials exposed to different solutions at different concentrations and temperatures. The corrosion tables showed that at any ammonium sulfate concentration, the corrosion rates of 304L, 316L and 317L is lower than 0.1 mm/year when the temperature is between 20 oC and the boiling point.   Moreover, Sandvik, which is a well-known Swedish company manufacturing stainless steel products among others, has also published its own corrosion tables in the public domain. Their laboratory corrosion tests were performed with pure chemicals and water almost saturated with air. They have also concluded that at any ammonium sulfate concentration, the corrosion rates of 3R12, 3R60, and 18Cr13Ni3Mo, which are equivalent to 304L, 316L and 317L, respectively, are lower than 0.1 mm/year when the temperature is between 20 oC and the boiling point.  3.2.3 Pitting Corrosion in Potable and Sea Water It is critical to review the literature regarding pitting corrosion in potable and seawater. This is required since the pipes were exposed to water during the hydrotesting and with improper draining and drying following the testing, residue water was stagnant for an extended period. Also, based on the background info, it was not clear which type of water was used. Therefore, it is important to briefly review the literature on this subject.   21  There are three main factors that should be considered when studying pitting corrosion in potable and seawater systems. These factors are chloride concentrations, temperature, and velocity. As for the chloride concentrations, it was briefly highlighted earlier that 316L stainless steels could be used for a range of 200 – 1000 ppm at around ambient temperature (Grubb, 2005). However, at a higher temperature, the required chloride concentrations to cause chloride induced pitting corrosion or chloride Induced Stress Corrosion Cracking (Cl SCC) is lower (Grubb, 2005). It is worth noting that for Cl SCC, the industrial practice is that no practical minimum limit of chloride is required to initiate such a damage mechanism. This is due to the well-established knowledge that there are circumstances where chloride concentrates such as during alternating exposure to dry-wet conditions. When this happens, Cl SCC could occur at elevated temperatures exceeding 140 oF (60 oC) (API 571, 2011).    Moreover, the second factor, temperature, should be considered and analyzed prior to recommending materials in seawater applications. Similar to the critical pitting temperature (CPT), critical crevice corrosion temperature (CCT) must be studied and measured using ASTM G 48 – 11 “Standard Test Methods for Pitting and Crevice Corrosion Resistance of Stainless Steels and Related Alloys by Use of Ferric Chloride Solution”. Such test methods are used to compare and rank the different types of stainless steels and their resistance to pitting and crevice corrosion. (Kovack, 1997)  22  As for the third factor, velocity, it is one of the most critical factors that should not be overlooked during the design, construction and commissioning stages. It was reported that at stagnant, intermittent or low water flow i.e. the velocity is less than 1.5 m/s, stainless steels would develop localized corrosion (Grubb, 2005). One of the causes for the development of localized corrosion is the formation of a microbial slime layer that causes biofouling (John Grubb, 2005). Another reason is the precipitation of deposits due to low/stagnant flow which creates Under Deposit Corrosion (UDC).  3.3    Microbiologically Influenced Corrosion (MIC) Microbiologically Influenced Corrosion (MIC) was first officially recognized by the National Association of Corrosion Engineers (NACE) to be the designated terminology to describe this type of failure (Javaherdashti, 2008). Yet, in literature, various terms were used to describe it, such as microbiologically induced corrosion, microbial corrosion, biocorrosion and biodeterioration. It may seem more appropriate to address it as Biologically Influenced Corrosion since the culprit here is not only microorganisms but also algae that can be seen by the unaided-eye which can have detrimental effect resulting in under-deposit corrosion. However, it is advisable to follow NACE and the literature norm. So, in this thesis, this type of corrosion is referred to as Microbiologically Influenced Corrosion (MIC).  23  3.3.1 Susceptible Materials to MIC MIC attack was reported to be the cause of failure in many types of materials. However, greater focus was directed towards materials that are extensively used in industrial applications. Therefore, it is reported that iron based alloys, copper based alloys, nickel based alloys and aluminum based alloys were among the most reported, studied and investigated materials in regard to their susceptibility to MIC. Moreover, in general, studies show that titanium and titanium based alloys are almost immune. (Jack, 2002)  Filip found that some types of bacteria even attack plastics (Filip, 1992). This highlights the fallacy in the common impression that non-metallic materials are superior substitutes to metallic materials.   Javaherdashti explained that the widespread reports of MIC could be a result of some kind of a chemical response (Javaherdashti, 2008). However, it seems that trying to correlate the MIC attack to some materials is not a practical approach since focusing the research to understand the attachment of the planktonic cells to any material’s surface to be sessile cells and the following microbial colonization is more logical. In other words, if the bacteria are present in an environment that provides nutrients to colonization growth, then the surface underneath this slime or biofilm will provide a localized environment or a concentration cell that is significantly different from the bulk stream, irrespective of the chemical composition of the base material. As such, the different bacteria that have oxidizing, reducing or producing mechanisms, provide a wide range of localized corrosive 24  media that attack different types of materials. Such an explanation is supported by Groysman (2010) who stated that the surface material has almost no effect on the biofilm attachment and growth since microorganisms attach to different metallic and non-metallic materials within 30 seconds of exposure with almost the same rate (Groysman, 2010). The next section highlights in detail the specific damage mechanisms of different types of bacteria.   3.3.2 Mechanisms of MIC Bacteria Types During the research for industrial case studies where the damage mechanism was attributed to MIC, many stipulated that Sulfate Reducing Bacteria (SRB) or Iron Oxidizing Bacteria (IOB) were the responsible types of bacteria. However, as highlighted by Geissler in the paper “Do not Just Blame the SRBs and APBs for MIC”, it stated the long common belief of blaming the bacteria which generate hydrogen sulfide (H2S) or organic acids as part of their metabolism to be responsible for MIC attacks. However, it is not only these two types that could cause MIC (Geissler, 2015). A NACE publication about the selection, application and evaluation of biocides in the oil and gas industry lists eight (8) different types of bacteria in oilfield environments: acid producers, iron oxidizers, iron reducers, general heterotrophs, methanogens, nitrate reducers, sulfur oxidizers, and sulfur reducers (NACE, 2006). The reason behind the mainstream focus on the two aforementioned types of bacteria to be studied, analyzed and linked to MIC was the fact that laboratory culturing of such types are easier than the rest (Geissler, 2015).   25  Nevertheless, in this thesis literature review, the focus was on studying the mechanisms of Sulfate Reducing Bacteria (SRB) and Iron Oxidizing Bacteria (IOB). This is due to the significant similarities between this failure in question and other failures that occurred in the industry with similar environment and materials of construction.    3.3.2.1 Sulfate Reducing Bacteria (SRB) Sulfate Reducing Bacteria (SRB) were comprehensively and frequently studied and analyzed to understand their corrosive effect on many metallic materials under different environmental conditions. Yet, other groups like thiosulfate-reducing bacteria, nitrate-reducing bacteria, acetogenic bacteria and methanogenic archaea were also suspected to be influencing the corrosion behaviour of iron and iron based alloys. Nevertheless, SRB was normally identified as the primary cause of failures attributed to MIC. This conception was based on the following three aspects, as highlighted and quoted below by Enning (2014):  “First, anoxic sulfate-rich environments (e.g., anoxic seawater) are particularly corrosive.”  “Second, SRB and their characteristic corrosion product iron sulfide are ubiquitously associated with anaerobic corrosion damage.”  “Third, no other physiological group produces comparably severe corrosion damage in laboratory-grown pure cultures.” (Enning, 2014) 26  Therefore, a hypothesis of MIC SRB corrosion should be corroborated by comparing the SRB mechanism and the other previously mentioned types of bacteria’s mechanism based on these three aspects.   For SRB to take place, there are two different scenarios that should be highlighted. These scenarios are Chemical Microbiologically Influenced Corrosion (CMIC) and Electrical Microbiologically Influenced Corrosion (EMIC). The former starts with the attack on iron by hydrogen sulfide which is the CMIC. The latter is when sulfate reducing bacteria attack the iron by the direct utilization of the metal itself which happens by the direct electron uptake, i.e. EMIC. It is worth mentioning that such scenario where noticed only on few strains of SRB that were studied in the last few years (Enning, 2014).  Enning in an earlier publication assumed that EMIC is wide spread and has significant technical relevance (Enning, 2012). Thus, CMIC and EMIC are considered the main processes for the corrosion of iron by sulfate reducing bacteria (Enning, 2014). Nevertheless, SRB attack can be further aggravated when oxygen ingresses in the system (Lee, 1995). Oxygen ingress could result in the formation of corrosive sulfur species due to the partial oxidation of dissolved hydrogen sulfide and iron sulfide deposits at the surface of the steel component (Nielsen, 1993).   27  3.3.2.2 Iron Oxidizing Bacteria (IOB) Iron oxidizing bacteria, or iron bacteria for short, utilize Fe(II) as an energy source, which oxidizes ferrous iron to ferric iron. The main feature of iron oxidizing bacteria like Gallionella is their vital part in concentrating the chloride ions under the tubercules because of the reaction of the chloride ions with the produced ferric ions by the iron oxidizing bacteria. As such, the localized environment under the tubercules consequently become very acidic. (Javaherdashti, 2013)  3.3.3 Morphology of the MIC Attack The morphology of MIC attack is one of the most discussed, studied, argued and also debated in the academic and industrial research communities. Some authors confirmed their failure investigation to be due to MIC solely based on the morphology of the attack. Others have supported their hypothesis of MIC failure by conducting bacteria cell count or other methods to analyze microbial presence and growth. Such methods are briefly described in a later section.  Borenstein (2002) analyzed a failure case of austenitic stainless steel pipe in contact with untreated stagnant water for 9 months. The through-wall pitting was sectioned and they found a large subsurface cavity with a small opening from the surface. In order to ensure that this type of attack was due to nothing but MIC, the authors used low chloride water since the presence of chloride will most likely trigger a localized attack in the form of pitting or crevice corrosion. (Borenstein, 2002) 28  It should be noted that the morphology of the MIC attack might be different when comparing carbon steel to stainless steels. In a laboratory investigation of carbon steel failed by MIC due to hydrotest water remaining in the pipes for an extended period of time, the pit morphology was different. Unlike the narrow opening and tunneling, pits caused by MIC in carbon steel are rather shallow and large. Sometimes exhibiting a morphology called cup-shaped pits. (Huang, 2012)  It is worth mentioning that the above two references are not the only ones. Such observations became well accepted identification of MIC pits such that many industrial standards, guidelines and reports highlighting MIC morphology of the attack in stainless steel is a wide subsurface cavity with a narrow opening whereas for carbon steel is cup-shaped pits within pits. This was highlighted in the American Petroleum Institute Recommended Practice (API 571) titled “Damage Mechanisms Affecting Fixed Equipment in the Refining Industry”.   3.3.4 Locations of MIC Failures Recognizing the probable or potential locations of MIC attacks are extremely beneficial to understand and readily identify any future damage due to MIC in order to better prevent or minimize their criticality and adverse consequences. As such, this section highlights the abundant industrial and academic studies about MIC attack locations.   29  Prior to reviewing the literature on this subject, it is imperative to define the word “weldment”. The term “weldment” includes the weld area, the heat affected zone (HAZ) and the adjunct parent metal, as noted by NACE/ASTM G193 “Standards Terminology and Acronyms Relating to Corrosion”. Therefore, with this in mind, this section reviews the publications based on the aforementioned definition. (NACE/ASTM, 2012)  Davis in his book “Corrosion of Weldments” explained that MIC usually occurs at or near the weld area and the heat affected zone. Kobrin in his article “Microbiologically Influenced Corrosion of Stainless Steels by Water Used for Cooling and The Hydrostatic Testing” identified that for weld defects like lack of penetration and root concavity, such defects are normally the preferred sites for attachment and growth of microbial colonies. (Kobrin, 1997)  As a general concept, MIC is not the only damage mechanism that attacks stainless steels at the weld area. Cramer and Covino have listed other damage mechanisms such as pitting corrosion, crevice corrosion, sensitization, Chloride Stress Corrosion Cracking (Cl SCC) and caustic embrittlement. There are many causes of the stainless steel weldment being the target of many damage mechanisms. Such reasons are:  During the welding process, the material exhibits a heating and cooling cycle which consequently affects the weldment microstructure and chemical composition. In this case, the corrosion resistance of non-filler welds will be lower than of the base metal. This is also true when using filler material for 30  welding that is the same in chemical composition to the base metal. (Cramer, 2003)  The selected heat input and the welder workmanship will change how the weldment solidifies, which, if done improperly, may reduce the corrosion resistance. (Streicher, 1978)  The corrosion of stainless steel weldments based on the aforementioned causes could be abated through the proper material selection of the base metal, by quality monitoring the welding practices and workmanship, and by using a suitable filler material. (Cramer, 2003)  Furthermore, Planktonic bacteria are presumably attracted to the weld area, fusion line or the heat affected zone (HAZ). Therefore, improper welding practices increase the chances of MIC initiation either by introducing heat oxide tent or by root concavity or lack of weld filler penetration which result in welding defect and a preferred zone for bacteria to attach to the surface and form the biofilm layer that later causes the MIC. (Cramer, 2003)  As highlighted earlier, it is critical to shed some light on why microorganisms prefer to attach themselves and colonize in the weldment area. One explanation for this issue is that due to welding, the surface roughness and even the chemical composition are different in a way that attracts the bacteria, or microorganisms in general, to colonize the 31  pipe internal surface that is in contact with water. Another explanation suggests that even if the weld area and base metal have the same chemical composition and there is no welding defect, bacteria tend to also attach to the weldment area because of residual stresses introduced by the welding. (Cramer, 2003)  3.3.5 Effect of Water Quality on MIC The quality of water which is used in hydrostatic testing is critical to the prevention of MIC at the equipment and piping commissioning stages. To ensure the suitability of water for the intended hydrostatic testing, a number of parameters should be controlled including chloride content, bacteria count, oxygen concentration, and pH. Therefore, this section discusses, in brief, the two main parameters: chloride concentration and the bacteria count.   In general, the variations of the previously mentioned parameters could be explained by the type of water used and the subsequent treatment that was applied, if implemented. For example, potable water does have lower chloride concentration and bacteria count than seawater. As for the aquifer and well water, it depends on the geographical location, as in near industrial or marine environment, and the depth of the wells. Depending on the type of water used in hydrotesting, the severity of potential MIC could be correlated.   Dexter (2003) stressed that the first encounter of material to the possible microorganisms is during the hydrostatic testing right after the piping fabrication in the shop. He continued 32  that, usually, the type of water used is untreated fresh well waters that have a considerable amount of bacteria, like Gallionella, which is blamed for causing MIC attacks on stainless steels. Therefore, the best line of defense is to treat the type of water used for hydrotesting by biocides. (Dexter, 2003)  As for which type of biocides should be used to prevent the microbial presence in the water, the oxidizing biocides are generally considered for such use. In particular, chlorine and ozone are the most chemicals used in the cooling water systems (Rice, 1991). However, due to regulation pertaining to the use of chlorine, ozone is increasingly used over it since only minimal rates of dosing are required compared to chlorine (Videla, 1995). As a common industrial practice highlighted by Javeherdashti, it was stipulated to control the bacteria count to 1000 cells/mL prior to using the water for hydrotesting purposes. As such, the water is classified as a low corrosive type (Javeherdashti, 2013).    33  4 OBJECTIVES  Based on the review of the literature and the presented background information about the failure of 316L SS pipes in ammonium sulfate solution in a premature manner, the objectives in this thesis are:  Conduct electrochemical tests i.e. open circuit potential and potentiodynamic polarization tests on 316L Stainless Steel electrodes in ammonium sulfate solution at various chloride concentrations and different ammonium sulfate concentrations. The objective of these tests is to prove that the material is suitable under conditions that are similar to the service of the failed pipes and to add evidence to the theory that the corrosion occurred before the pipes were put into service.     Confirm that the pipe material meets the nominal chemical composition of 316L stainless steel and that it is not mechanically different from standard 316L.   Perform a detailed failure analysis.  34  5 ELECTROCHEMICAL TESTING ON 316L SS IN AMMONIUM SULFATE SOLUTION Several electrochemical tests in ammonium sulfate solution were conducted to simulate the actual conditions of the process of the failed piping.  Open Circuit Potential (OCP) and Potentiodynamic Polarization (PDP) were conducted at different ammonium sulfate concentrations and various chloride concentrations. The purpose was to evaluate if the ammonium sulfate solution which was transported inside the piping was corrosive to the piping material of construction, 316L SS. Thus, this chapter includes the testing method and the discussion of the results on 316L SS electrodes.   5.1 Specimen Preparation  A cylindrical sample was cut into coin-shape from a 316L SS rod. A Teflon insulated wire was attached to the back of the sample using conductive epoxy. The sample was then mounted in cold-cure epoxy.  Grinding was performed to the mounted sample to remove first the thin layer of epoxy, that sometimes exist due to imperfect handling of the resin, in order to have a uniform and planar surface. Then, further grinding/polishing was done to 1200 grit silicon carbide paper to remove any course scratches. Polishing was then done with 6 um and then 1 um diamond suspension. Finally, the specimen was washed with methanol, rinsed with demineralized water and then dried with an air jet.   35  5.2 Electrolyte Composition  Ammonium sulfate solution (NH4)2SO4 was prepared using white granular (99% min.) Ammonium sulfate ACS grade and demineralized water. Sodium chloride (NaCl) was added to the solution at different concentrations. Table 5.1 show the prepared and tested electrolyte composition. Note that all the tests conducted in naturally aerated solutions at room temperature.   Table 65.1: Testing Environment and Conditions  Testing Material Conducted Electrochemical Tests Testing Conditions (NH4)2SO4 NaCl 1 316L SS Open Circuit Potential and Potentiondynamic Polarization 100 g/l 2 – 4 g/l (at 1 g/l increment) 2 316L SS Open Circuit Potential and Potentiondynamic Polarization 100 g/l 6 – 8 g/l (at 1 g/l increment) 3 316L SS Open Circuit Potential and Potentiondynamic Polarization 400 g/l 0 and 1 g/l   5.3 Corrosion Cell Setup A standard three-electrode corrosion cell was used to conduct the electrochemical tests. A 1-liter of the ammonium sulfate solution was added to the cell followed by immersing the prepared sample of 316L SS as the working electrode opposite to the counter electrode made from graphite. Also, a silver chloride (Ag-AgCl) electrode was installed 36  acting as a reference electrode. A “VersaSTAT-4” potentiostat/galvanostat of Princeton Applied Science was used to perform the open circuit potential and the potentiodynamic polarization tests in order to analyze the corrosion behaviour of the 316L SS in the intended service.   5.4 Electrochemical Techniques    5.4.1 Open Circuit Potential (OCP) The OCP, sometimes called the corrosion potential, is defined as the potential difference of the working electrode compared to the reference electrode when no current is applied. The OCP tests of ammonium sulfate solution of different concentrations, 100 g/l and 400 g/l, and various chloride concentrations, 0 to 10 g/l of NaCl, were carried out at room temperature. The measured values of the voltage in reference to the Silver-Silver chloride electrode were plotted against the time in seconds.   5.4.2 Potentiodynamic Polarization Test The method of potentiodynamic polarization is to change the working electrode potential and measure the corresponding current as a function of potential/voltage. Based on the test findings, the corrosion current density can be determined. The potentiodynamic polarization tests were conducted following the OCP tests under the same conditions as highlighted in the previous Table 5.1. Only anodic polarization tests were conducted. The scan rate was set at 1 mV per 6 seconds i.e. 0.167 mV/s. Additionally, at exactly the same concentrations and testing parameters, two anodic polarization tests on 316L SS electrode were conducted while only varying the scan rate from 0.167 mV/s to 1 mV/s.  37  5.5 Results and Discussion of the Electrochemical Tests on 316L SS Electrode  5.5.1 Effect of Increasing Chloride Concentration Figures 5.1 – 5.3 illustrate the OCP test results for the 316L Stainless Steel working electrode in naturally aerated ammonium sulfate solution at pH of 5.4 with various chloride concentrations as a function of time. Figure 5.1 illustrates the noticeable increase in the open circuit potential when increasing the chloride content from 2 g/l to 4 g/l in 10% ammonium sulfate solution. Figure 5.2 tells a slightly different story when the chloride content was increased from 6 g/l to 8 g/l while maintaining the other parameters at the same values. Although the measured potential at the beginning of the test showed a significant difference between the different chloride concentrations, when the time passed and reached the 3600 s mark, the open circuit potential of 316L stainless steel for the different chloride concentrations almost converged. Looking closer, the open circuit potentials increased at a millivolt (mV) level with the increase of chloride content. As for Figure 5.3, the open circuit potentials in 40% ammonium sulfate solution were measured to compare the effect of adding 1 g/l of chloride with the same solution without the presence of chloride. The results showed that the OCP increased with the addition of chloride, which is consistent with the previous OCP tests.    38   Figure 45.1: Open Circuit Potential of 316L SS in 2 – 4 g/l of chloride in 100 g/l ammonium sulfate solution    Figure 55.2: Open Circuit Potential of 316L SS in 6 – 8 g/l of chloride in 100 g/l ammonium sulfate solution  -0.3-0.2-0.100.10.20.30.40.50 500 1000 1500 2000 2500 3000 3500 4000Open Circuit Potential (V vs. Ag/AgCl Electrode)Time (s)2 g/l NaCl3 g/l NaCl4 g/l NaCl-0.2-0.15-0.1-0.0500.050.10.150 500 1000 1500 2000 2500 3000 3500 4000Open Circuit Potential (V vs. Ag/AgCl Electrode)Time (s)6 g/l NaCl7 g/l NaCl8 g/l NaCl39   Figure 65.3: Open Circuit Potential of 316L SS in 400 g/l ammonium sulfate solution  Figures 5.4 – 5.6 illustrate the potentiodynamic polarization results on 316L SS while increasing the chloride concentration. Figure 5.4 shows that in 10 wt.% of ammonium sulfate solution, increasing the chloride from 2 g/l to 4 g/l causes a noticeable shift to higher corrosion potentials and corrosion current densities. In Figure 5.5, with the same ammonium sulfate concentration, further increasing the chloride content from 6 g/l to 8 g/l increased the corrosion current densities on the 316L SS electrode. Moreover, at 400 g/l of ammonium sulfate solution, Figure 5.6, two tests were conducted to see the effect of a chloride-free solution and chloridized solution with 1 g/l NaCl. The results illustrated an increase in the corrosion potentials and corrosion current densities while adding the chloride to the solution. These tests showed that with the addition of chloride in the solution, the corrosion current density increases. Specifically, the corrosion current -0.1-0.0500.050.10.150.20.250 1000 2000 3000 4000Open Circuit Potential (V vs. Ag/AgCl Electrode)Time (s)No Chloride Added1 g/l NaCl40  density for the 316L SS electrode in chloridized solution with 1 g/l NaCl was 1.74 µA/cm2. Therefore, the corrosion rate was 0.8 mpy i.e. 0.02 mm/year. This very low corrosion rate agrees with the previously conducted studies highlighted in the literature review chapter.   Figure 75.4: The effect of increasing chloride Content from 2 - 4 g/l in 100 g/l ammonium sulfate on the 316L working electrode  -0.6-0.4-0.200.20.40.60.811.21.41.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04Potential (V vs. Silver Chloride Electrode)Current Density (A/cm2)100 + 2100 + 3100 + 441   Figure 85.5: The effect of increasing chloride content from 6 - 8 g/l in 100 g/l ammonium sulfate on the 316L working electrode  Figure 95.6: The effect of presence of the chloride ions on the corrosion behaviour of the 316L working electrode in 400 g/l ammonium sulfate solution -0.4-0.200.20.40.60.811.21.41.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03Potential (V vs. Ag/AgCl Electrode)Current Density (A/cm2)100 + 6100 + 7100 + 8-0.500.511.51.00E-10 1.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03 1.00E-02Potential (V vs. Ag/AgCl Electrode)Current Density (A/cm2)0 g/l of Cl1 g/l of Cl42  5.5.2 Effect of Increasing Ammonium Sulfate Concentration  OCP experiments to measure the effect of the increase of ammonium sulfate concentration rather than the chloride concentration on 316L SS were also conducted. Figure 5.7 shows the open circuit potential of 316L SS in chloridised ammonium sulfate solution at 1 g/l of NaCl. When the ammonium sulfate concentration increased from 10% to 40%, the open circuit potential was significantly increased. The potentiodynamic polarization tests, Figure 5.8, showed that when increasing the ammonium sulfate concentration from 100 g/l to 400 g/l in the test solution, the corrosion potentials and the corrosion current densities increased.   Figure 105.7: Open Circuit Potential of 316L SS in chloridised ammonium sulfate solution at 1 g/l of NaCl  -0.25-0.2-0.15-0.1-0.0500.050.10.150 500 1000 1500 2000 2500 3000 3500 4000Open Circuit Potential (V vs. Ag/AgCl Electrode)Time (s)400 g/l AS100 g/l AS43   Figure 115.8: The effect of increasing the ammonium sulfate concentration from 100 to 400 g/l on the 316L working electrode  5.5.3 Effect of Increasing the Scan Rate of the Anodic Polarization Test At exactly the same concentrations and testing parameters, two anodic polarization tests on 316L SS electrode were carried out while only varying the scan rate from 0.167 mV/s to 1 mV/s. The ammonium sulfate concentration was 100 g/l with 5 g/l of NaCl. Figure 5.9 illustrates the effect of increasing the scan rate six times. The test results shows that increasing the scan rate altered the measured results as the passive film had more time to form at the slower scan rate. In an industrial setting, the passive film would have ample time to form so the corrosion rate shown here (0.8 mpy) may actually be an overestimate.  -0.6-0.4-0.200.20.40.60.811.21.41.61.00E-10 1.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03Potential (V vs. Ag/AgCl Electrode)Current Density (A/cm2)100 g/l AS with 1 g/l NaCl400 g/l AS with 1 g/l NaCl44   Figure 125.9: Anodic Polarization Curves show the effect of increasing the scan rate in the experiment on the measured potentials and current densities.    -0.4-0.200.20.40.60.811.21.41.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03Potential (V vs. Ag/AgCl Electrode)Current Density (A/cm2)Slow Scan RateHigh Scan Rate45  6 TESTING PROCEDURE FOR THE FAILED 316L SS PIPES Conducting a proper investigation of the failed equipment usually results in significant information about the critical factors that led to the failure. This wealth of knowledge aids in avoiding the recurrence of the failure. Hence, in order to properly organize a thorough failure analysis, it is essential to have a basic understanding of the various procedural steps which are utilized during the metallurgical laboratory investigation. Therefore, in this chapter, an in-depth explanation is presented for the conducted tests on the failed pipes through metallographic, chemical and mechanical testing analyses.   6.1 Visual Examination  One of the first and foremost steps in any metallurgical failure investigation is the visual examination. Visual examination can help to identify, in broader terms, the fractured/damaged surface characteristics and any visible surface defects.  During this step, no destructive test is used. All of the following tests were conducted at the sponsoring Company analytical laboratories.  Although dye penetrant test (PT) is considered a non-destructive testing (NDT) method and is usually conducted during the visual examination step of the failed sample, such a test was not conducted during this examination due to two reasons. First, PT is normally used to clearly show all the locations of pits and surface cracks, which is not necessary for this failure at hand since the failure and pits are visible by the unaided eye. The second reason is that the PT may contaminate the surface if an EDS analysis, as part of the 46  scanning electron microscope, is carried out in an as-received condition, which is the case in our investigation.   Therefore, prior to destructive sampling, many photographic records of key observations were taken. Such records are referred to as macrophotographs, which usually utilize magnifications between 1x and 50x. These photographs were taken using digital camera and stereomicroscope, i.e. a low-power binocular optical microscope, with the proper light sources.   6.2 Chemical Testing  6.2.1 Chemical Analysis of the Process Sample Three samples from the ammonium sulfate process were obtained from the same unit of the failed pipes. The samples were analyzed to determine the pH and the chloride contents. The chloride concentration in the samples was analyzed using Ion Chromatography, Dionex model DX-500. The analysis was conducted at the facilities of the sponsoring Company.     6.2.2 X-Ray Fluorescence and Carbon/Sulfur Analyses One of the essential starting points in the metallurgical investigation of the failures is the identification and confirmation of the chemical composition of the failed sample. As in the case of the failure at hand, the background information showed that the material of 47  construction was in fact 316L stainless steel. Therefore, X-Ray Fluorescence (XRF) and Carbon/Sulfur analyses were conducted. All of the following tests were conducted at the sponsoring Company analytical laboratories.  XRF is an analytical technique which is used for solid samples to illustrate the concentrations of the elements. XRF spectrometers operate by irradiating the solid sample with an X-ray beam of a high energy in order to excite the characteristic X-rays from the elements that are present in the solid sample. In this investigation, the Bruker S4 Pioneer XRF spectrometer was utilized.   For Carbon/Sulfur analysis, the LECO CS-600 Automated Analyzer was utilized. This analyzer is designed for the prompt determination of carbon and sulfur in various materials such as ferrous and non-ferrous alloys, and some non-metallic materials.   The analyzer operation concept is straightforward. Since the majority of metals and their alloys burns in oxygen when heated to significantly elevated temperatures, the carbon oxidizes to carbon dioxide (CO2) and sulfur converts to sulfur dioxide (SO2). The installed Infrared detectors then measure the CO2 and SO2. Moreover, in order to increase the speed and the accuracy, high-frequency induction heated furnaces are used.    48  6.3 Mechanical Testing In order to find if the failed pipe meets the stipulated mechanical properties as per ASTM A312 “Standard Specification for Seamless, Welded, and Heavily Cold Worked Austenitic Stainless Steel Pipes”, it is necessary to conduct the required mechanical testing. Therefore, this section covers the most commonly used mechanical tests: tensile test and micro-hardness test. All of the following tests were conducted at the sponsoring Company analytical laboratories.   6.3.1 Tensile Tests Tensile testing is utilized to supplement microscopy and confirm its results in the investigation of the components’ failures. The tensile test indicates the level of the ductility and the strength.   Two pipe samples were cut and prepared to determine the tensile strength and elongation. The specimens’ dimensions were as stipulated in the Figure 6.1 and Table 6.1 below. 49   Figure 136.1: The Tensile Test Sample. Note the actual dimensions in Table 6.1.  Table 76.1: The Dimensions of the Specimens Used in the Tensile Tests  Parameters Measured Values 1 The Width of Reduced Section 12.5 mm 2 Thickness 3.0 mm 3 Gauge Length 50.0 mm 4 The Width of Grip Section 20.0 mm 5 Grip Section  50.0 mm 6 Overall Length 200.0 mm  The tensile tests were conducted at two rates. The first rate purpose was to identify the yield properties where the strain rate of 0.015 mm/mm/min up to 2.0% of tensile strain. The second rate was used to identify the tensile properties where the strain rate of 0.402 mm/mm/min up until the complete and final fracture.   The tensile tests were conducted using a 50kN Instron Universal Tester at room temperature of 21.7°C. The strain was calculated by 50 mm extensometer up to 2% strain 50  followed by Linear Variable Differential Transformer (LVDT) of the test machine. In the next chapter, Results and Discussion, the test results are presented and discussed.  6.3.2 Hardness Tests Hardness testing is one of the fastest, most effective and inexpensive tests used in failure investigations. Hardness is the measure of the resistance of a material to the indentation. Although there are several types of hardness test, Vickers Microhardness test was utilized in this investigation. The main reason behind selecting this type is due to its very narrow indenter which determines the localized hardness measurements and differentiates between local hard spots and the surrounding area.    The purpose of a microhardness test is to determine the hardness of the material in very localized areas. As in this current investigation, the microhardness tests were conducted across the pipe cross-sectional area which includes: weld area, Heat Affected Zone (HAZ) and base metal.    The principles governing the hardness testing is basically the same whether Rockwell Testing or Vickers Microhardness was used.  An indenter is pressed into the material by known weights on a level system. The depth of penetration and the covered diamond-shaped area are then measured and the hardness value of the material can be calculated. 51  The Vickers indentor is a standard square-based diamond pyramid with 136o angle α between opposite faces and 146o angle between opposite edge of the pyramid. The type of the utilized machine is Buehler MMT 7 Microhardness Tester.   The tests were conducted at room temperature (18.4°C) and the load was 300 gf. The tests were conducted according to the following methods: 1. Standard Guide for Preparation of Metallographic Specimens – ASTM E3 2. Standard Test Method for Microindentation Hardness of Materials – ASTM E384  6.4 Metallurgical Testing 6.4.1 Samples Preparation for Metallography Tests As stated previously, the first step in the metallographic analysis is to select a sample that is representative of the materials to be tested. Then, the second step is to prepare the metallographic specimens. Generally, five major operations are involved in preparing the metallographic specimens: sectioning, mounting, grinding, polishing, etching.  It is vital to carefully implement such operations in order to ensure that the tested specimens were not altered due to improper preparation practices. Incorrect preparation techniques may negatively affect the microstructure and could result in inaccurate findings which points to an incorrect cause of failure.   52  After sectioning and grinding, polishing was conducted. The goal of polishing is to remove the scratches and the heavily deformed layer that was caused by grinding. Polishing is the last step for metallographic specimens’ preparation prior to the microscopic examination. After polishing, the specimen is cleaned by solvents for two purposes: examining the polished surface using optical microscope followed by SEM/EDS and etching the polished surface in order to examine under the optical microscope in as-etched condition.    Therefore, in order to analyze the failed samples at hand, two samples were cut and mounted by compression molding, followed by grit silicon carbide grinding from 120 µm to 600 µm. Successively, the samples surfaces’ were polished using diamond abrasive (6 µm).  6.4.2 Stereoscope and Optical Microscope Analyses Optical metallography is the examination of materials using visible light to illustrate magnified images of the micro and macrostructures. Microscopy is basically the microstructural examination at magnifications of approximately 50x or higher. However, macroscopy, i.e. macro-structural examination or simply stereoscope, involves magnifications of 50x or lower. Microscopy and Macroscopy can both be utilized for as received, as polished or as etched specimen evaluation.   53  6.4.3 Scanning Electron Microscopy and Energy-Dispersive X-Ray Spectroscopy Analyses SEM/EDS was used for the characterization of the metal structures and chemical compositions, as well as in analyzing the fracture surfaces.  An Oxford INCA 250 EDS installed on a Quanta 200 SEM system was utilized for the analysis of the failure. Magnifications from 25x to 2000x were employed in the analysis of multiple specimens. Both types of specimens, i.e. the mounted and polished specimens and the as-received samples were analyzed. The as-received samples without polishing or even cleaning were studied intensively for two purposes: to illustrate the damaged surface morphology and to analyze the chemical composition of any deposits or matters that may be discovered during the process. Utilizing the capabilities of SEM, the surface morphology was revealed in order to identify defects’ characteristics due to the high resolution and depth of field of the SEM.  It is important to note that the x-ray peak energies for different elements might overlap. Such an issue might lead to incorrect identification of elements, especially for 316L SS. An example of peak overlap is between molybdenum and sulfur. Since 316L SS contain 2-3% of Mo, the EDS quantitative analysis might show sulfur around this weight percentage. Missing this critical understanding of the peak overlaps might derail the investigation to totally different direction by assuming high amount of sulfur present in the material, for example. Another common peak overlaps are for Mo and Nb, Ti and V, Cr and Mn, and Cr and oxygen. See Annex B of ISO 22309 “Microbeam Analysis - 54  Quantitative Analysis using Energy-Dispersive Spectrometry (EDS) for Elements with an Atomic Number of 11 (Na) or Above”. (ISO 22309, 2011)   55  7 RESULTS AND DISCUSSION FOR THE FAILED 316L SS PIPES: FAILURE ANALYSIS In this chapter, all data and information derived from the failure analysis and laboratory tests are presented. Moreover, a detailed analysis and careful interpretation of the test results is described.  7.1 Visual Examination  Two samples from the 3-inch diameter failed pipe (spools) where received for the metallurgical examination. A general overview of the received spools is shown in Figures 7.1 & 7.2. Meshed-like marking was observed covering part of the external surface of the spools. The pinhole leak location is almost at the center of the meshed-like marking. After inquiring from the plant inspectors, they highlighted that such markings are due to quick fixes, i.e. using metallic clamps with rubber pads, in order to stop the leak while the line was in operation to avoid interrupting the plant production by isolating the damaged pipe which requires an unplanned shutdown. 56   Figure 147.1: Photograph of the failed spool in as received condition. Note that the pinhole located at the weldment area.  Figure 157.2: Photograph of the other failed spool in as-received condition. Note that the pinhole here is located at the parent metal.  Pinhole Location Tensile Test Sample Tensile Test Sample 57  Although the pinholes located at or near the 6 o’clock position in both spools, the vicinity of the pinholes are not consistent. The pinhole in one of the spools is located near the weldment while the other is in the parent metal, Figures 7.3 & 7.4. Such observation would indicate that such failures are not only due to welding defects. On the other hand, this observation may also indicate that the causes of the pinholes might not be the same. 58   Figure 167.3: A Close-up view of the failed spool showing the pinhole at the fusion line.  Figure 177.4: A Close-up view of the failed spool showing the pinhole at the parent metal. 2.8 cm pipe section was cut - see Fig. 12 2.6 cm pipe section was cut - see Fig. 10 59  Figure 7.5 shows the internal surface of the failed spool. The close-up view illustrates discoloration and staining on the surface. The white deposits are from the ammonium sulfate solution.   Figure 187.5: Close-up view of the internal surface of the failed spool.  Furthermore, a closer view of the lower portion of the spool indicated an interesting finding, Figures 7.6 & 7.7. The spool was cut near the pinhole location and a subsurface cavity, sometimes called tunneling, was observed. The tunneling indication coupled with the extended idle period of 3 years where water was stagnant suggest that the probable 60  cause of failure is Microbiological Influenced Corrosion “MIC” as documented in many published reports.    Figure 197.6: Photograph shows the pinhole and the subsurface tunneling after sectioning the spool circumferentially. Note the close-up views of the pinhole mouth at the ID. 61   Figure 207.7: A Close-up view of the sectioned spool near the pinhole.   Moreover, the visual examination of the internal surface showed the morphology of the weld, see Figures 7.8 & 7.9. A closer view of the weld illustrated weld defects like a lack of penetration and root concavity. This is an indication of poor weld quality which could be attributed to improper welding procedure and/or poor workmanship. 62   Figure 217.8: Photograph shows the internal surface of the failed spool.    Figure 227.9: Close-up view shows the multiple pits initiated at the incomplete root penetration area of the weld as indicated in Figure 7.8. Incomplete Root Penetration (Weld Defect) Complete Root Penetration (Proper Weld) Pits Pits at the weld - see Fig. 7.9 63  Another observation from the visual examination is related to the origin of the pinholes. It is essential to determine whether the pinholes initiated from the internal surface or the external surface. A closer view of both surfaces highlighted that there are multiple separated pinholes found on the inner surface but only one from the outer side. Such finding indicates that the pitting was initiated from the inner surface.  Finally, the photographs clearly illustrated multiple pits located few millimeters away from the weld, in the Heat Affected Zone (HAZ). Also, other pits located at the weld in the area that was affected by lack of penetration. Davis (2006) explained that MIC usually occurs at or near the weld and the heat affected zone. This investigation revealed that most of the pinhole leaks were located at the weld and HAZ. Kobrin, in his article “Microbiologically Influenced Corrosion of Stainless Steels by Water Used for Cooling and The Hydrostatic Testing” identified that for weld defects like lack of penetration and root concavity, such defects are normally the preferred sites for attachment and growth of microbial colonies. This was clearly illustrated as the pits located in such areas in Figure 23. (Davis, 2006; Kobrin, 1997)   64  7.2 Chemical Testing 7.2.1  Chemical Analysis of the Process Samples Three samples from the ammonium sulfate process were obtained from the same unit of the failed pipes. The samples were analyzed to determine the pH and the chloride contents. Table 7.1 below highlights the findings.   Table 87.1: pH and chloride concentrations in 3 process samples Parameter Unit Sample 1 Sample 2 Sample 3 pH --- 3.0 3.80 3.2 Chloride ppm 0.06 0.03 0.15  These results illustrate that chloride is in a very low amount the in the ammonium sulfate process. When compared with the electrochemical tests that I have conducted in UBC corrosion lab, the conducted tests were at 10,000 ppm of chloride. Moreover, even at 10,000 ppm of Cl, the 316L stainless steel is adequate for the service. Therefore, the ammonium sulfate process with low chloride content did not cause the corrosion failure at hand.     7.2.2   X-Ray Fluorescence and Carbon/Sulfur Analyses The provided failed samples were analyzed using XRF and C/S analyses. The results confirmed that the material of the pipe is SS 316L, see Table 7.2 below. Note that the carbon and sulfur were analyzed using LECO CS-600 Automated Analyzer.  65    Table 97.2: XRF and carbon-sulfur analysis for the two failed pipes  Sample # 1 Sample # 2 Nominal Composition wt. % Elements Conc. % Conc. % Max. unless otherwise indicated C* 0.019 0.018 0.035 Al 0.033 0.045  Si 0.528 0.474 1.00 P 0.016 0.019 0.045 S* 0.016 0.016 0.030 Ti 0.006 0.018  V 0.059 0.058  Cr 17.236 17.313 16.0 – 18.0 Mn 1.381 1.555 2.00 Fe Balance Balance  Ni 10.294 10.200 10.0 – 14.0  Cu 0.111 0.112  Nb 0.008 0.008  Mo 2.074 2.110 2.00 – 3.00     7.3 Mechanical Testing 7.3.1 Tensile Tests As stipulated in the previous chapter regarding the steps to properly conduct the tensile testing, Figure 7.10 and Table 7.3 illustrate the test results. The two samples both show that the tensile strength, yield strength, and elongation are all exceeding the minimum requirements. Thus, the material of the examined spool is in compliance with the mechanical properties stipulated in ASTM A312 Type 316L. Therefore, it is safe to assume that the mechanical properties of the pipe are not compromised and that the material in itself did not contribute to the failure at hand.     66   Figure 237.10: Tensile test results show the tensile and yield strength of the two samples.   Table 107.3: Tensile tests results illustrate that both samples meet the mechanical properties requirements for the type 316L SS. Grade UNS Designation Tensile  strength, min MPa Yield  strength, min MPa Elongation in 2 in. or 50 mm (or 4D), min, % TP316L S31603 485 170 35 Sample 1 - 645 405 54.9 Sample 2 - 640 407 53.4  7.3.2 Hardness Tests Microhardness test was conducted and the results are stated in Table 7.4 and Figure 7.11. The results of the microhardness tests showed values ranging between 169 and 187 HV. These values are not exceeding the maximum requirement of the hardness 67  values of austenitic stainless steel type 316L as stipulated in ASM International Handbook “Minimum Room-Temperature Mechanical Properties of 304L and 316L Austenitic Stainless Steels” (ASM International Handbook, 1990). The maximum hardness values of 316L SS is 95 HRB, which is equivalent to 213 HV. As stated previously, all the hardness values are below the maximum hardness value of the pipes.  Table 117.4: Microhardness values for the failed sample including the weld area, HAZ and parent metal.  Readings, µm  Indentation No.  D1 D2 Hardness (HV) 8 @ 1 mm Interval 54.2 55.1 186.2 7 54.2 55.5 184.9 6 53.8 55.1 187.6 5 54.1 55.5 185.2 4 57 57.6 169.4 3 57.2 56.5 172.1 2 56.2 55.7 177.7 1 56.7 57 172.1 0 (Center of the Weld) 57.1 57 170.9 1 55.7 55.6 179.6 2 (Pinhole Area) -- -- -- 3 56 55.5 178.9 4 56.3 56.6 174.5 5 55.3 56.8 177 6 55.5 55.1 181.9 7 55.8 55.2 180.6  68   Figure 247.11: Photomacrograph showing the exact location of the indentations for the microhardness testing.  7.4 Metallurgical Testing  7.4.1 Stereoscope and Optical Microscope Analyses Two samples were cut from the failed spools and prepared through grinding, mounting, and polishing to be examined using the optical microscope. Figures 7.12 -7.14 showed a large subsurface cavity with only a small opening to the internal surface. This finding indicates that the pinhole was originally initiated from the internal surface. Also, it supports the evidence that this failure is due to MIC. As it was stated in the literature review chapter, the pit morphology caused by MIC tends to be a narrow opening with a wide subsurface cavity and tunneling at the pits sites. Although this is not a definite way to identify MIC as pits with similar morphology could be formed by other damage mechanisms as well, the 100110120130140150160170180190200-9 -7 -5 -3 -1 1 3 5 7 9Hardness (HV)Location in the Weldment (mm)Hardness Readings Across the Weldment69  metallurgical findings along with the piping commissioning history suggest that the damage mechanism here is Microbiologically Influenced Corrosion (MIC). Further tests such as biological and chemical analyses could be conducted to strengthen this conclusion. However, the objective here is to focus on the metallurgical examination and pinpoint to MIC failures based on the conducted tests in this regard.     Figure 257.12: Photograph shows a prepared and hot mounted sample of the localized damage at the weld area.  OD ID 70   Figure 267.13: Photomicrograph shows the localized attack in as polished-condition.  Figure 277.14: Photomicrograph shows the localized attack in as etched-condition. 71  Another sample which was prepared for metallography was selected away from the location of the pits to examine the microstructure of the base material. Figures 7.15 & 7.16 show the photomicrographs in as-polished and as-etched conditions, respectively. The photomicrographs indicated normal microstructure without any abnormal inclusions. Such findings indicate that the base metal is in sound condition and did not contribute to this failure.    Figure 287.15: Photomicrograph of the 316L SS spool sample away from the pinhole in as-polished condition. 72   Figure 297.16: Photomicrograph of the 316L SS sample away from the pinhole in as-etched condition.  7.4.2 Scanning Electron Microscopy and Energy-Dispersive X-Ray Spectroscopy Analyses A Scanning Electron Microscope equipped with EDS elemental analyzer was utilized in order to identify the morphology of the pitting in as received condition as well as the cross-sectional profile through the pit.   The sample in Figure 7.8 with the pinhole located in the weldment was further investigated using SEM. No cleaning was performed prior to placing the sample inside the SEM chamber in order to avoid losing any vital evidence that helps further in the investigation. 73  Thus, Figure 7.17 shows the cross-section photograph of the subsurface cavity and tunneling at 20x magnification. At higher magnifications, 70x, 100x, and 250x, as in Figures 7.18, 7.19, and 7.20, respectively, the SEM photographs show an interesting damage called preferential attack mechanism in the weld metal precipitates. As highlighted by Jack (2002), the morphology of the MIC attack in stainless steel weldments is preferential phase attack where he described it as “interdendritic attack”, see Figure 7.21. Such description fits the examined photomicrograph in my analysis of the failure under SEM at the fusion line. This type of failure is usually related to preferential attacks on either ferrite or austenite. Figures 7.18 - 7.20 illustrate the interdendritic attack at the fusion line of the stainless steel weldment which show similar to Jack analyzed failure.  Figure 307.17: SEM image at low magnification showing the cross-section photograph of the subsurface cavity and tunneling. 20x Magnification.  74   Figure 317.18: SEM image showing the cross-section photograph of the subsurface cavity. Note the preferential attack mechanism. 70x Magnification  Figure 327.19: SEM image showing the cross-section photograph of the subsurface cavity. Note the preferential attack mechanism. 100x Magnification 75   Figure 337.20: SEM photomicrographs showing the cross-section of the subsurface cavity. Note the preferential attack mechanism. 250x Magnification A B 76   Figure 347.21 has been removed due to copyright restrictions. It was a picture that shows the morphology of MIC attack on the fusion line of SS weldment. Jack compared between the dendritic attack to the nondendrite area. Original source: Thomas R. Jack, Biological Corrosion Failures, ASM Handbook, Volume 11, 2002.  This preferential attack was also found during the analysis of the internal surface of the other failed sample in as-received condition. Figure 7.22, illustrates a photograph of a pit that is located at the fusion line. At higher magnification, Figures 7.23 and 7.24 show again the preferential attack in the direction of the rolling.  77   Figure 357.22: SEM image shows the pinhole in as received condition. Magnification 25x.  Figure 367.23: A Magnified SEM image of the same location of Figure 7.22. Note the dimension of the pits openings. Magnification 100x. Fusion Line Weld Area HAZ 78   Figure 377.24: SEM image shows the same previous pinhole at higher magnification (200x). Note the clear preferential attack mechanism at the edge of the pit.  Furthermore, the second sample was analyzed using EDS to reveal the elemental analysis at the critical locations. Figure 7.25 shows a high magnification of the same location as in Figure 7.20-A. As shown in Figure 7.25-B, two spot analyses, S1 and S2, were conducted at a distinctive feature. The EDS results are demonstrated in Figures 7.27 & 7.28 and Tables 7.5 & 7.6. One unanticipated finding was that the analysis of spot S1 revealed a high amount of carbon around 60%, while S2 spot shows a lower amount of Carbon. It is encouraging to compare this Figure with that found by Dexter who highlighted that this is the shape of bacteria, Figure 7.26 (Dexter, 2003). Therefore, the high carbon content coupled with the distinct shape suggest the presence of bacteria that could cause this failure. While such a shape could simply be atmospheric carbon-containing dust, the shape and location provide some evidence that this is bacterial remnants. 79    Figure 387.25: SEM images of the cross-section of the pinhole in as-received condition. Note the irregular shape suggesting bacteria presence. Mag. 1200x and 2000x. A B S1 S2 80   Figure 397.26 has been removed due to copyright restrictions. It was a photomicrograph of a form of bacterial cells in a colony at a magnification of 2700x. Original source: Stephen C. Dexter, Microbiologically Influenced Corrosion, ASM Handbook, Volume 13A, ASM International, 2003.    Figure 407.27: EDS analysis for the S1 spot in Figure 7.25. Note the high Carbon peak.  Table 127.5: Qualitative EDS analysis for the S1 spot. Note the high Carbon content suggesting the presence of bacteria. Element Wt% At% C 59.92 74.10 O 21.50 19.96 Si 00.36 00.19 Mo 01.42 00.22 Ca 09.57 03.55 Cr 03.10 00.89 Fe 04.13 01.10 Matrix Correction ZAF 81   Figure 417.28: EDS analysis for the S2 spot in Figure 7.25. Notice the high Iron peak and the low Carbon peak.  Table 137.6: Qualitative EDS analysis for the S2 spot. Element Wt% At% C 08.75 24.27 O 14.57 30.35 Si 00.58 00.69 Mo 06.26 02.18 Cr 20.94 13.42 Fe 46.12 27.52 Ni 02.78 01.58 Matrix Correction ZAF    82  Figure 7.30 and Table 7.7 illustrate the EDS analysis of a deposit on the internal surface of the pipe as shown in Figure 7.29. The elemental analysis showed high sulfur content in the deposit. The presence of sulfur-containing deposit suggests that the type of bacteria involved in the Microbiologically Influenced Corrosion (MIC) is Sulfate Reducing Bacteria (SRB). As discussed in the literature chapter, MIC failures in the stainless steel pipes were commonly attributed to SRB.    Figure 427.29: SEM Image shows deposits on the internal surface of the pipe. 83   Figure 437.30: EDS analysis of the deposits as shown in Figure 7.29.  Table 147.7: Qualitative EDS analysis for deposit Area. Element Wt% At%   N 21.23 29.15   O 41.91 50.39  Al 00.72 00.51  Si 00.61 00.42   S 28.51 17.11  Fe 07.02 02.42 Matrix Correction ZAF  The conducted metallurgical analysis tests showed that the failure was caused by Microbiologically Influenced Corrosion (MIC) based on the pit morphology, the preferential attack mechanism in the weld and the presence of bacteria/bacteria product.  Yet, it could be argued that localized attack in this form of pitting could also result from 84  chloride induced pitting corrosion. Such argument can be contested by analyzing the samples to look for any chloride concentration present in the as-received samples. The EDS analysis illustrated in Figures 41 and 42 did not show any indication of chloride presences. However, these were spot elemental analyses at high magnification at around 2000x. So, in order to clearly refute this argument, multiple EDS area analyses were conducted for the location illustrated in Figure 7.20, which is at a lower magnification i.e. around 250x. As shown in Figures 7.31 & 7.32 and Tables 7.8 & 7.9, no chloride was detected in the elemental analysis by EDS. Therefore, this finding suggests that chloride did not play a role in the failure at hand. In comparison, for failures caused by chloride ions, as reported by Subai (2014), they detected chloride using EDS at the bottom surface of the pipe at lower magnification in different locations at 4.16 %wt., 2.19 %wt. and 0.14 %wt. (Subai, 2014). Also, they cross-sectioned the pit and conducted EDS analysis in as-polished condition and found 4.76%wt. of chloride at the bottom surface of the pit (Subai, 2014). Based on Subai study, if the failure at hand was due to chloride induced pitting corrosion, then the conducted EDS analysis should detect it.  85   Figure 447.31: EDS analysis (Above) for the pinhole in as-received condition (Below).  86  Table 157.8: Qualitative EDS analysis for the area specified in Figure 7.31. Element Wt% At% C 05.92 18.12 O 11.11 25.54 Si 01.00 01.31 Mo 03.77 01.45 S 01.65 01.90 Cr 27.94 19.76 Fe 45.62 30.05 Ni 02.99 01.87 Matrix Correction ZAF  Table 167.9: Qualitative EDS analysis for the area specified in the following Figure 7.32. Element Wt% At% C 11.94 32.97 O 09.09 18.84 Si 01.51 01.78 Mo 03.14 01.09 S 00.53 00.55 Cr 23.47 14.97 Fe 47.30 28.09 Ni 03.03 01.71 Matrix Correction ZAF  87   Figure 457.32: EDS analysis (Above) for the pinhole in another location in as-received condition (Below).  88  7.5   Effect of Welding on MIC Initiation As pointed out in the background chapter of this thesis, the conducted radiographic testing (RT) showed rounded and elongated indications which were observed in the weld joints for the full circumference. This non-destructive testing results suggest improper welding performed for the pipes joints. These indications could be caused by inadequate Welding Procedure Specification (WPS) or poor workmanship which is more likely. These defects adversely affect the pipe joint integrity since the material will be susceptible to MIC at the weldment.   As previously noted in the literature review, planktonic bacteria are presumably attracted to the weld area, fusion line or the heat affected zone (HAZ). Therefore, improper welding practices increase the chances of MIC initiation either by introducing heat oxide tent or by root concavity or lack of weld filler penetration which result in welding defect and a preferred zone for bacteria to attach to the surface and form the biofilm layer that later causes the MIC.   As highlighted earlier, it is critical to shed some light on why microorganisms prefer to attach themselves and colonize in the weldment area. One explanation for this issue is that due to welding, the surface roughness and even the chemical composition are different in a way that attracts the bacteria, or microorganisms in general, to colonize the pipe internal surface that is in contact with water. Another explanation suggests that even if the weld area and base metal have the same chemical composition and there is no 89  welding defect, bacteria tend to also attach in the weldment area because of residual stresses introduced by the welding. (Cramer, 2003)  7.6   Effect of Hydrostatic testing procedure on MIC Initiation As described earlier in the background chapter, there are improper practices occurred during the construction which could adversely affect the integrity of the piping. This section discussed the potential issues of the water quality that might be used during the construction phase for the hydrotesting.   It is important to know which type of water was used and how it was treated. The current investigation was limited in this regard. The plant inspectors claimed that the used water was in compliance with their plant standards. Such standards stipulated the use of demineralized water, boiler condensate water, or potable water that was treated effectively to limit the chloride and oxygen concentrations and the total bacteria count. However, they failed to provide any documentation supporting their claim. Therefore, with the failure at hand coupled with the lack of proof, the used water for hydrotest was at best untreated or poorly treated potable water. A possible explanation is that during the construction phase, the project inspectors may delegate the responsibility of the hydrotesting to the contractor with little or even no quality monitoring or supervision from their side. Such practices are not unusual in large project construction with tight deadlines.    90  Therefore, it is likely that the water had a high bacterial count due to poor water treatment. With this in mind, and in order to reduce the possibility of MIC initiation, the water should have a bacteria count of less than 1000 (Javeherdashti, 2013). This can be achieved by using biocide treatment to effectively reduce the planktonic and therefore sessile bacteria in the system.  7.7   Effect of the Idle period between construction and commissioning on MIC propagation  This section discusses one of the main contributing factors in this failure, i.e. the extended contact time of the hydrotest water inside the piping. It is basically the idle time between conducting the hydrotesting as a final phase of the construction and the time of commissioning the new plant.   As discussed in the background chapter, the idle time lasts for 3 years. This is alarming because constructing the piping, hydrotesting it and then leaving it in closed condition without proper lay-up makes the pipeline prone to different undesirable outcomes.   In the literature review chapter, many references were cited about the time limit upon which the MIC start to initiate and/or propagate. The cited intervals are as follows: 3 to 5 days, 14 days and 1 months. If the idle time planned to be higher than such interval, then dry or even wet lay-up should be conducted. Nevertheless, the failure at hand crossed all these intervals and reached more than 1000 days. Yet, they have not protected their 91  assets by conducting preservation method like purging with an inert gas in order to prevent the piping from corrosion.   Simply put, the more time the bacteria are living in their ideal environment, the more bacteria growth is expected, assuming the availability of the nutrient in the environment. In general, when the used water for hydrotesting is less clean, such as having high total dissolved solids (TDS), the more nutrient is available in the water for bacteria to consume and grow.  It is worth mentioning that during the investigation of the failure at hand, other piping in the same system which was constructed a year earlier were also experiencing pinholes. Although these new failures were not investigated thoroughly, it is safe to assume that the longer the idle period where water is in contact with the steel surface, the more pinholes and deeper pits will be developed. This is based on that the time from commissioning the pipes to the observed leaking pinhole is different. For the pipes with 3-year idle period, the first leak was noticed after only 4 months in service. However, for the 1-year idle period, the first leak was noticed after 8 months. The other failures with the one-year idle period were not included in this investigation.     92  8 CONCLUSIONS AND RECOMMENDATIONS  8.1 Summary of the Findings from the Conducted Laboratory Tests In this investigation, one of the main objectives was to identify the responsible damage mechanism that consequently caused this failure and address, in details, all the contributed factors that played significant roles in this failure. The investigation of the pinhole defects in the 316L austenitic stainless steel pipes in ammonium sulfate solutions suggested that the failure was caused by Microbiologically Influenced Corrosion (MIC). The following highlights the main conclusions:  1. The Low Corrosivity of Chloridized Ammonium Sulfate Solution to 316L SS The conducted electrochemical tests illustrated the low corrosivity of the chloridized ammonium sulfate solution to 316L stainless steel. The corrosion rate based on the anodic polarization test of 316L SS electrode in chloridized solution with 1 g/l NaCl would not have exceeded 0.02 mm/year. This conclusion is supported by other studies conducted on 316L stainless steel in chloridized ammonium sulfate at a higher temperature reaching 100 oC. These studies and the published industrial corrosion tables concluded the 316L SS is suitable for ammonium sulfate solution at any concentration with a corrosion rate less than 0.1 mm/year.     2. Pitting Morphology The most obvious finding to emerge from this investigation is the classical pits morphology of MIC. It was numerously reported that a pit with a small opening “mouth” 93  and a large subsurface cavity with tunneling in stainless steels is likely to be a “signature” of an attack by MIC. The visual and optical microscope examinations revealed the pitting morphology that is similar to dozens of reported industrial cases with MIC identified as the damage mechanism.   3. Pitting Location Another important finding of this investigation is the location of the pinhole and the other pits. It was evident that almost all the pits are located at or near the fusion line, i.e. the heat affected zone (HAZ). This result was explained that the microorganisms prefer to attach to the surface of the weldment. Such attachments could be due to different surface roughness or different chemical composition between the weld area and the base metal. Also, the attachment might be due to welding defects or residual stresses from the welding.   4. Preferential Attack Mechanism  One of the most interesting findings is the preferential attack mechanism. The SEM photomicrographs showed the preferential attack of ferrite stringers at the fusion line which resulted in a dendritic surface at the pit. It was repeatedly highlighted in the literature that the morphology of the MIC attacks in stainless steel weldments is a preferential attack of a single phase.    94  5. The Presence of Bacteria Surprisingly, one of the unanticipated findings during the SEM examination was that bacteria may have been identified. There were two indications that suggest the existence of bacteria inside the pit. The first indication was the shape. When compared with a Figure by Dexter (2003), it illustrated the same shape of a bacterial cell in a colony. The second indication was the elemental analysis. The EDS elemental spot analysis at the bacterial cell showed a spike of carbon whereas in the adjacent area, the spot analysis did not reveal the high carbon content.   Such finding strengthens the hypothesis that the MIC is the damage mechanism that led to the piping multiple failures. As mentioned in the results and discussion chapter, the analyzed sample was in as-received condition. Moreover, the pipe was sectioned through the pit using dry cutting to reveal the subsurface cavity and tunnelling. Therefore, no contamination was expected from the external environment.     6. The Absence of chloride  Generally, when pitting occur in austenitic stainless steel materials, it is widely believed that chloride might be the main cause of the corrosion attack. Therefore, during the investigation, the metallurgical examination of the failed sample in the as-received condition did not show the presence of chloride. Thus, the absence of chloride means that the failure was not due to chloride induced pitting corrosion which in turns strengthen the conclusion that the failure occurred due to MIC considering the other vital findings.  95  8.2 Summary of Key Observations from the Investigation If we take one step away from all the detailed findings which were revealed by the electrochemical, metallurgical, chemical and mechanical tests, the bigger picture could be visualized. As such, MIC attack could not occur without the below deficiencies that paved the way to the failure.  1. Improper Hydrotesting Procedure As highlighted in the background, the plant inspectors did not provide documentation that proof the completion of all the required steps of the hydrotesting and the following draining and drying. The residual water after the hydrotest was not properly drained and dried. Therefore, a thorough draining and drying followed by effective inspection are critical to ensure the dryness of the pipes immediately after the hydrotest which help to prevent MIC in the first place.   2. The Welding Quality  The investigation revealed that the examined weld was performed improperly. Visual examination showed welding defects in the form of root concavity and lack of penetration. Such welding defects may favor the planktonic bacteria to attach to these areas and colonize. Therefore, ensuring the proper workmanship and avoiding any welding defects will safeguard the integrity of the pipe joint and reduce the likelihood for planktonic to be sessile. 96  3. Idle Time “Contact Time” A 3-year idle time between hydrotesting the piping during the construction phase and the commissioning is very alarming. It was reported that a maximum of 30 days should be between these two phases. If the period is expected to exceed this 30-day interval, the line should be mothballed by purging inert gas to avoid active corrosion. However, the plant professional reported that after hydrotesting, the line was closed and purging with an inert gas was simply not performed.  8.3 Recommendations to Avoid the Reoccurrence of the Failure Based on all the presented findings from the metallurgical examination and the highlighted observations from the investigation in general, below highlights the main recommendations: 1.  It is recommended to replace the failed pipes with the in-kind material, i.e. 316L Stainless Steel. The failure of the pipes was not attributed to improper material selection.  2. It is strongly recommended to use treated water for the hydrostatic testing. Using seawater, well water or even untreated potable water which have high bacteria count will increase the possibility of damaging the pipes by Microbiologically Influenced Corrosion (MIC). 3. It is recommended to completely drain and dry the pipes after the hydrotesting. Failure to do so will result in water accumulation and stagnation which will increase the likelihood of MIC attack on the pipes.  97  4. It is recommended to minimize the idle period between the construction of the pipes and the commissioning of the unit to be preferably less than 14 days. If the anticipated idle period will be longer than a month, then the pipes should be preserved and protected by purging the pipes internally with nitrogen.    8.4 Suggested Future Work  This thesis presented and discussed in greater details the metallurgical investigation of the failed 316L SS pipes, which includes determining the causes of failures and the associated contributing factors that led to promptly end the useful life in a premature manner. Nevertheless, such work could be expanded in future to include conducting the following:  1. Electrochemical Testing: Although the thesis discussed and studied several aspects of the failure at hand including conducted OCO and Anodic Polarization tests, it deemed beneficial to expand the conducting electrochemical tests to study the corrosivity of the ammonium sulfate solution at simulated conditions to the industrial plant that experienced the failure. The simulated tests should include the addition of Urea at concentrations of 0.2 wt.% with a pH value of 3.0. Also, the tests should include measuring the corrosion current density at different solution states: stagnant and flowing conditions. Such study will highlight if in a stagnant condition the 316L SS material would experience higher corrosion rate and by what magnitude.  98  Moreover, besides ammonium sulfate solution, the electrochemical tests could also be conducted for different types of water, i.e. seawater, well water and potable water. The objective here is to identify the corrosion rate of each type of water in order to possibly narrow-down the water type which was used in the hydrotest that resulted in such failure.   2. MIC Testing This thesis investigated the failure based on the metallurgical analysis. However, such investigation could be expanded to include conducting various microbial tests that are usually used in MIC laboratory investigations. Such tests could include analyzing different types of water and measure the bacteria count and the types of bacteria present in the analyzed samples. The analysis could also include treating the water samples with different types of biocides at different concentrations to measure what would be the more effective method for MIC prevention.   Such MIC testing could also be coupled with chemical analysis of the water samples. Analyzing the concentrations of TDS, TSS, chloride, oxygen, and pH, in relation to the bacteria count and the performance of the biocides could be deemed valuable for establishing significant and fundamental correlations.   99  BIBLIOGRAPHY  K. Alawadhi, M. Martinez, A. Aloraier, J. Alsarraf, Effect of inhibitors on preferential weld corrosion under sweet conditions, Anti-Corrosion Methods and Materials. 61, 172-181, 2014. P. E. Arnvig and A. D. Bisgard, “Determining the potential independent critical pitting temperature (CPT) by a potentionstatic method using the avesta cell”, NACE International Conference, 1996.  American Petroleum Institute, Recommended Practice API RP 571, Damage Mechanisms Affecting Fixed Equipment in the Refining Industry, 2nd edition, 2011.  ASM International Handbook Committee, ASM Handbook, Volume 01 - Properties and Selection: Irons, Steels, and High-Performance Alloys, ASM International, 1990. ASTM, Standard # A312, Standard Specification for Seamless, Welded, and Heavily Cold Worked Austenitic Stainless Steel Pipes, 2017. ASTM, Standard # E3, Standard Guide for Preparation of Metallographic Specimens, 2011. ASTM, Standard # E384, Standard Test Method for Microindentation Hardness of Materials, 2017. R.E. Avery, S. Lamb, C.A. Powell, A.H. Tuthill, Stainless steel for potable water treatment plants (PWTP)-Guidelines, Nickel Development Institute, 1-8, 1999. 100  Tiziano Bellezze, Giampaolo Giuliani, Gabriella Roventi, Study of stainless steels corrosion in a strong acid mixture. Part 1: cyclic potentiodynamic polarization curves examined by means of an analytical method, Corrosion Science, Volume 130, 113-125, 2018. M.V. Biezma, Microstructural features and the microbiologically influenced corrosion of stainless steels, European Federation of Corrosion Publications (UK), 29, 36-46, 2000. P.E. Frederick Bloetscher, R.J. Bullock, G.M. Witt, Brackish Water Supply Corrosion Control Issues using 3161 Stainless Steel, 2001. S.W. Borenstein, P.B. Lindsay, Microbiologically influenced corrosion failure analysis of 304L stainless steel piping system left stagnant after hydrotesting with city water, NACE Corrosion Conference, 2002. J.E. Bringas, Handbook of Comparative World Steel Standard, ASTM International, 2007. L. Carpén, T. Ohligschläger, A Case Study On Corrosion Of Stainless Steel In Firewater Mains, 6th European Stainless Steel Conference, Finland, 2008. C. Chandrasatheesh, J. Jayapriya, R.P. George, U.K. Mudali, Detection and analysis of microbiologically influenced corrosion of 316 L stainless steel with electrochemical noise technique, Eng. Failure Anal. 42, 133-142, 2014. J. H. Cleland, What Does the Pitting Resistance Equivalent Really Tell Us?, Engineering Failure Analysis, Vol. 3, No. 1, 65-69, 1996. 101  T.E. Cloete, Biofouling control in industrial water systems: what we know and what we need to know, Materials and Corrosion. 54, 520-526, 2003. S.D. Cramer, B.S. Covino Jr, ASM Handbook, Volume 13B - Corrosion: Materials, ASM International, 2005. S.D. Cramer, B.S. Covino Jr, ASM Handbook, Volume 13A - Corrosion: Fundamentals, Testing and Protection, ASM International, 2003. Y. Cui, C.D. Lundin, Austenite-preferential corrosion attack in 316 austenitic stainless steel weld metals, Materials & design. 28, 324-328, 2007. A. Darwin, K. Annadorai, K. Heidersbach, Prevention of Corrosion in Carbon Steel Pipelines Containing Hydrotest Water-An Overview, NACE Corrosion Conference, 2010. J.R. Davis, Corrosion of Austenitic Stainless Steel Weldments, in: Corrosion of Weldments, ASM International, 43-57, 2006. M.Y. Demeri, Advanced High-Strength Steels - Science, Technology, and Applications, ASM International, 2013. Stephen C. Dexter, Microbiologically Influenced Corrosion, ASM Handbook, Volume 13A, ASM International, 2003. James Dillon, P. Desch, T. Lay, The Nalco Guide to Cooling Water Systems Failure Analysis, McGraw-Hill Education, 2nd edition, 2015. The Editor, Chloride and pH limits for stainless steel, Mater. Perform. 49, 59, 2010. 102  F. Elshawesh, K. Abusowa, H. Mahfud, A. Abderraheem, F. Eljweli, K. Zyada, Microbial-Influenced Corrosion (MIC) on an 18 in. API 5L X52 Trunkline, Journal of Failure Analysis and Prevention. 8, 60-68, 2008. Dennis Enning, Julia Garrelfs, Corrosion of Iron by Sulfate-Reducing Bacteria: New Views of an Old Problem, Applied and Environmental Microbiology, Volume 80, 1226-1236, 2014.  D. Enning, H. Venzlaff, J. Garrelfs, H. Dinh, V. Meyer, K. Mayrhofer, A. Hassel, M. Stratmann, F. Widdel. Marine Sulfate-Reducing Bacteria Cause Serious Corrosion of Iron Under Electroconductive Biogenic Mineral Crust. Environ. Microbiol, 1772-1787, 2012.  G.S. Frankel, Microbiologically Induced Corrosion, ASM Handbook, Volume 13 A, 2003. R.P. George, P. Muraleedharan, K.R. Sreekumari, H.S. Khatak, Influence of surface characteristics and microstructure on adhesion of bacterial cells onto a type 304 stainless steel, Biofouling 19, 1-8, 2003.  R.P. George, P. Muraleedharan, Microbiologically Influenced Corrosion, Corrosion of Austenitic Stainless Steels, Mechanism, Mitigation and Monitoring, 218-240, 2002.  Alec Groysman, Corrosion for Everybody, 2010. A Ghahremaninezhad, W Wang, E Asselin, J Budac, A Alfantazi, Electrochemical and corrosion behaviour of stainless steels 316L and 317L in chloridised ammonium sulfate solution, Canadian Metallurgical Quarterly, 51:4, 471-484, 2012. 103  John F. Grubb, Terry DeBold, James D. Fritz, Corrosion of Wrought Stainless Steels, ASM Handbook, Volume 13B, Corrosion: Materials, ASM International, 54-75, 2005.   E. Hargrave, Behavior of 300-Series Stainless Steel Heat Exchangers in Cooling Water Service, 2004. W. Huang, D. Xu, G. Ruschau, J. Wen, J. Hornemann, T. Gu, Laboratory Investigation of MIC Due to Hydrotest Using Seawater And Subsequent Exposure to Pipeline Fluids With And Without SRB Spiking, NACE Corrosion Conference, 2012.  E. Huttunen-Saarivirta, M. Honkanen, T. Lepistö, V.-. Kuokkala, L. Koivisto, C. Berg, Microbiologically influenced corrosion (MIC) in stainless steel heat exchanger, Applied Surface Science. 258, 6512-6526, 2012.  M.A.M. Ibrahim, El Rehim, S S Abd, M.M. Hamza, Corrosion behavior of some austenitic stainless steels in chloride environments, Mater. Chem. Phys. 115, 80-85, 2009. D. Ifezue, F. Tobins, Failure Investigation of a Gate Valve Eye Bolt Fracture During Hydrotesting, Journal of failure analysis and prevention. 13, 249-256, 2013.  ISO, Standard # 22309, Microbeam Analysis - Quantitative Analysis using Energy-Dispersive Spectrometry (EDS) for Elements with an Atomic Number of 11 (Na) or Above, 2011.   Thomas R. Jack, Biological Corrosion Failures, ASM Handbook, Volume 11, 2002.  R. Javaherdashti, Examples of Some Materials Vulnerable to MIC, Microbiologically Influenced Corrosion: An Engineering Insight, 123-131, 2008.  104  Reza Javaherdashti. Microbiologically Influenced Corrosion, Corrosion and Materials in Oil and Gas Industries, 47-128, 2013.  M.A. Javed, S.L. McArthur, P.R. Stoddart, S.A. Wade, Techniques for studying initial bacterial attachment and subsequent corrosion of metals, Corros. Preven. 2, 299-304, 2013.  S. Kaewkumsai, S. Auampan, K. Wongpinkaew, E. Viyanit, Root cause analysis for 316L stainless steel tube leakages, Eng. Failure Anal. 37, 53-63, 2014.  K.V. Kasiviswanathan, N.G. Muralidharan, N. Raghu, R.K. Dayal, H. Shaikh, Corrosion Related Failures of Austenitic Stainless Steel Components, in: Corrosion of Austenitic Stainless Steels, 314-339, 2011.  Y. Kikuchi, F. Matsuda, Microbiologically influenced corrosion of metal welds, Welding international. 9, 677-681, 1995. Y. Kikuchi, F. Matsuda, K. Tomoto, T. Okayama, T. Sakane, Role of micro-organism on the microbiologically influenced corrosion (MIC) susceptibility of stainless steel weld, Biodeterioration and Biodegradation. 133, 91-99. 1996. E.W. Klechka, Use of corrosion inhibitors on the trans Alaska pipeline, Materials performance. 3, 7-9, 2001. G. Kobrin, Microbiologically Influenced Corrosion of Stainless Steels by Water used for Cooling and Hydrostatic Testing, 58th  Annual International Water Conference, Pittsburgh, PA, 1997. 105  P.S. Korinko, S.H. Malene, Considerations for the weldability of types 304L and 316L stainless steel, Practical Failure Analysis. 1, 61-68, 2001. Curtis W. Kovach, James D. Redmond, High Performance Stainless Steels and Microbiologically Influenced Corrosion, Avesta Sheffield Corrosion Management and Application Engineering, 1997.   M. Krauss, R. Herzog, B. Scholtes, Influence of surface treatment on the pitting corrosion behaviour of high alloyed stainless steels in a chloride solution, Zeitschrift Fur Metallkunde. 92, 910-915, 2001.  O. Kvernvold, R. Johnson, T. Helgerson, Assessment of internal pipeline corrosion, Proceedings of the International Conference on Offshore Mechanics and Arctic Engineering, 409, 1992. W. Lee, Z. Lewandowski, P. Nielsen, W. Hamilton, Role of Sulfate-Reducing Bacteria in Corrosion of Mild Steel: A Review. Biofouling 8, 165-194, 1995. A. N. Lebedev, V. A. Blinova, Y. N. Dulepov, The corrosion of stainless steels under evaporator conditions during the concentration of ammonium sulfate solution, Chemical and Petroleum Engineering 10, 797, 1974. B.J. Little, J.S. Lee, Microbiologically influenced corrosion, Wiley, Hoboken, NJ, 2007.  B. Little, R. Ray, P. Wagner, E.D. Thomas, Microbiologically Influenced Corrosion of Corrosion Resistant Materials, Tri-Service Committee on Corrosion Proceedings, 1994. W. Liu, Rapid MIC attack on 2205 duplex stainless steel pipe in a yacht, Engineering Failure Analysis. 42, 109-120, 2014. 106  L.L. Machuca, L. Murray, R. Gubner, S.I. Bailey, Evaluation of the effects of seawater ingress into 316L lined pipes on corrosion performance, Materials and Corrosion, 65, 8-17, 2014. L.L. Machuca, S.I. Bailey, R. Gubner, E.L. Watkin, M.P. Ginige, A.H. Kaksonen, K. Heidersbach, Effect of oxygen and biofilms on crevice corrosion of UNS S31803 and UNS N08825 in natural seawater, Corrosion Science. 67, 242-255, 2013. H. Mansoori, R. Mirzaee, A. Mohammadi, Pitting Corrosion Failures of Natural Gas Transmission Pipelines, IPTC 2013: International Petroleum Technology Conference, 2013. T. Mathiesen, J. Frantsen, Unusual Corrosion Failures of Stainless Steel in Low Chloride Waters, NACE Corrosion Conference,  2008. O. McWane, Integrity of rewelded thin wall 316L stainless steel tube, Master's Theses 782, San Jose State University, 1994. A.S Modaihsh, Characteristics and Composition of the Falling Dust Sediments on Riyadh City, Saudi Arabia, Journal of Arid Environments, Volume 36, Issue 2, 211-223, 1997. D.A. Moreno, J.R. Ibars, J.L. Polo, J.M. Bastidas, EIS monitoring study of the early microbiologically influenced corrosion of AISI 304L stainless steel condenser tubes in freshwater, Journal of Solid State Electrochemistry. 18, 377-388, 2014. D.A. Moreno, J.R. Ibars, C. Ranninger, De Mele, M F L, H.A. Videla, Evaluation of Stainless Steel Susceptibility to Microbiologically Influenced Corrosion: An Updated 107  Review, Mycotoxins, Wood Decay, Plant Stress, Biocorrosion, and General Biodeterioration, Springer, 527–552, 1994. NACE International, Selection, Application and Evaluation of Biocides in the Oil and Gas Industry, NACE 31205 Technical Report, 2006. NACE/ASTM, G193, Standards Terminology and Acronyms Relating to Corrosion, 2012. R.C. Newman, Understanding the Corrosion of Stainless Steel, Corrosion. 57, 1030–1041, 2001. PH. Nielsen, W. Lee, Z. Lewandowski, M. Morison, WG. Characklis, Corrosion of Mild Steel in an Alternating Oxic and Anoxic Biofilm System. Biofouling 7, 267–284, 1993. M. Ohanian, V. Diaz, M. Corengia, P. Russi, M.J. Pianzzola, R.J. Menes, Microbiologically influenced corrosion evaluation on brass (UNS C68700, UNS C443) and AISI 316 stainless steel, REVISTA DE METALURGIA. 50, 2014. I. Olefjord, The passive state of stainless steels, Materials Science and Engineering. 42, 161-171, 1980. C.A. Olsson, D. Landolt, Passive films on stainless steels—chemistry, structure and growth, Electrochimica Acta. 48, 1093-1104, 2003. R. Ovarfort, “Critical Pitting Temperature Measurements of Stainless Steels with an Improved Electrochemical Method”, Corrosion Science, Vol. 29, 987-993, 1989. J.E. Penkala, J. Fichter, S. Ramachandran, Protection against microbiologically influenced corrosion by effective treatment and monitoring during hydrotest shut-in, NACE Corrosion Conference, 2010. 108  S.L. Percival, J.S. Knapp, D.S. Wales, R. Edyvean, Physical factors influencing bacterial fouling of type 304 and 316 stainless steels, British Corrosion Journal. 33, 121–129, 1998. Daniel E. Powell, Burkin Melancon, Robert H. Winters, Guidelines When Conducting Hydrostatic Pressure Tests and Management of Extended Dwell Times Through Dry/Wet Lay-Ups, NACE Corrosion Conference, 2011. R. Prasad, Chemical treatment options for hydrotest water to control corrosion and bacterial growth, NACE Corrosion Conference, 2003. B.S.C. Rao, R. Madeswaran, R. Chandramohan, In-fabrication and pre-service care on stainless steel pressure vessels, International journal of pressure vessels and piping. 73, 53–57, 1997. R.I. Ray, B.J. Little, J.S. Lee, The Role of Marine Bacteria in Stainless Steel Pitting, 2002. R.G. Rice, J.F. Wilkes, “Fundamental Aspects of Ozone Chemistry in Recirculating Cooling Water Systems,” NACE Corrosion Conference, Paper 205, 1991. M. H. W. Renner, "Corrosion Engineering Aspects Regarding MIC Related Failures on Stainless Steels", NACE Corrosion Conference, paper 285, 1998. J.J. Scutti, W.J. McBrine, ASM Handbook, Volume 11 - Introduction to Failure Analysis and Prevention, 2002. A. John Sedriks, Corrosion of Stainless Steels, 2nd edition, Wiley-Interscience, 1996. 109  H.M. Shalaby, A. Husain, Technical note: Localised corrosion of electric resistance welds of austenitic stainless steel in sea water, British Corrosion Journal. 27, 45–49, 1992. X. Shi, On the use of nanotechnology to manage steel corrosion, Recent Patents on Engineering. 4, 44–50, 2010. K. R. Sreekumari, K. Nandakumar, Y. Kikuchi, Bacterial attachment to stainless steel welds: significance of substratum microstructure, Biofouling. 17, 303–316, 2001. K.R. Sreekumari, Y. Sato, Y. Kikuchi, Antibacterial metals-A viable solution for bacterial attachment and microbiologically influenced corrosion, Materials transactions. 46, 1636–1645, 2005. D. Starosvetsky, O. Khaselev, J. Starosvetsky, R. Armon, J. Yahalom, The role of sulfides in iron activation in chloride-containing solutions, Electrochemical and Solid State Letters. 2, 265-266, 1999. J. Starosvetsky, D. Starosvetsky, B. Pokroy, T. Hilel, R. Armon, Electrochemical behaviour of stainless steels in media containing iron-oxidizing bacteria (IOB) by corrosion process modeling, Corrosion Science. 50, 540-547, 2008. J. Stewart, D.E. Williams, The Initiation of Pitting Corrosion on Austenitic Stainless-Steel - on the Role and Importance of Sulfide Inclusions, Corros. Sci. 33, 1992. J.G. Stoecker, Penetration of stainless steel following hydrostatic test, A practical manual on microbiologically influenced corrosion. Kobrin G (ed), NACE, Houston, Texas, 1993. 110  M.A. Streicher, Pitting Corrosion of 18cr-8ni Stainless Steel, J. Electrochem. Soc. 103, 375-390, 1956. M.A. Streicher, Theory and Application of Evaluation Tests for Detecting Susceptibility to Intergranular Attack in Stainless Steels and Related Alloys—Problems and Opportunities, Intergranulur Corrosion of Stainless Alloys, STP 656, American Society for Testing and Materials, 70, 1978. L.L.M. Suarez, Microbiologically Influenced Corrosion of Common Alloys Used for Subsea Applications, Citeseer, 2012. Saud Al Subai, Keshab Barai, Esteban Morales Murillo, Failure analysis of firewater pipeline, Case Studies in Engineering Failure Analysis, Volume 2, Issue 2, 144-149, 2014. M.E. Suess, Bacteria-induced corrosion of a stainless steel chemical trailer barrel, ASM International, Handbook of Case Histories in Failure Analysis. 1, 70–73, 1992. Kazuo Sugibayashi, Researches on the Corrosion of 18-8 Steel by the Aqueous Solution of Ammonium Sulphate, and its Prevention, the Society of Chemical Engineers, Japan, Volume 18, Issue 1, 16-21 1954. K. Sugimoto, Y. Sawada, Role of Alloyed Molybdenum in Austenitic Stainless-Steels in Inhibition of Pitting in Neutral Halide Solutions, Corrosion. 32, 347-352, 1976. T. Suter, H. Bohni, A new microelectrochemical method to study pit initiation on stainless steels, Electrochim. Acta. 42, 3275-3280, 1997. A.K. Tiller, Is stainless steel susceptible to microbial corrosion?, Microbial Corrosion, 104–107, 1983. 111  A.K. Tiller, Some case histories of corrosion failures induced by bacteria, International Biodeterioration. 24, 231–237, 1988. A.H. Tuthill, R.E. Avery, Survey of stainless steel performance in low chloride waters, Public Works. 125, 1994. R.N. Tuttle, Corrosion in oil and gas production, Journal of petroleum technology. 39, 756–762, 1987. J. E. Truman, Proceedings of UC Corrosion, Brighton, 111-129, 1987. B.R. Tzaneva, L.B. Fachikov, R.G. Raicheff, Effect of halide anions and temperature on initiation of pitting in Cr-Mn-N and Cr-Ni steels, Corrosion Engineering Science and Technology. 41, 62-66, 2006. T. Ujiro, K. Yoshioka, R.W. Staehle, Differences in Corrosion Behavior of Ferritic and Austenitic Stainless-Steels, Corrosion. 50, 953-962, 1994. M. Urgen, A.F. Cakir, The Effect of Molybdate Ions on the Temperature-Dependent Pitting Potential of Austenitic Stainless-Steels in Neutral Chloride Solutions, Corros. Sci. 32, 841-852, 1991. H.A. Videla, M.R. Viera, P.S. Guiamet, M.F.L. deMelle, J.C. Staibano Alais, Biocidal Activity of Dissolved Ozone on Sessile and Planktonic Bacteria: Effects of Corrosion Behavior of Cooling Water Systems Structural Materials, Proc. 1995 Intl. Conf. on MIC, P. Angel et al., Ed., NACE International, 62, 1995. D. Walsh, D. Pope, M. Danford, T. Huff, The effect of microstructure on microbiologically influenced corrosion, JOM. 45, 22–30, 1993. 112  T. Watanabe, K. Ei, H. Nakamura, Pitting Corrosion and Microscopically Measuring of the Local Corrosion Potential in Welds of Austenitic Stainless-Steels, Transactions of National Research Institute for Metals. 29, 87-97, 1987. Z. Wei, P. Duby, P. Somasundaran, Inhibition of pitting corrosion by surfactants as a function of temperature, Corrosion. 61, 341-347, 2005. D.E. Williams, J. Stewart, P.H. Balkwill, The Nucleation, Growth and Stability of Micropits in Stainless-Steel, Corros. Sci. 36, 1213-1235, 1994. S. Xin, J. Xu, F. Lang, M. Li, Effect of temperature and grain size on the corrosion behavior of 316L stainless steel in seawater, Materials and Manufacturing, Pts 1 and 2. 299-300, 175-178, 2011. C. Xu, Y. Zhang, G. Cheng, W. Zhu, Localized corrosion behavior of 316L stainless steel in the presence of sulfate-reducing and iron-oxidizing bacteria, Materials Science and Engineering: A. 443, 235-241, 2007. C. Xu, Y. Zhang, G. Cheng, W. Zhu, Pitting corrosion behavior of 316L stainless steel in the media of sulphate-reducing and iron-oxidizing bacteria, Materials Characterization. 59, 245-255, 2008. D. Xu, W. Huang, G. Ruschau, J. Hornemann, J. Wen, T. Gu, Laboratory investigation of MIC threat due to hydrotest using untreated seawater and subsequent exposure to pipeline fluids with and without SRB spiking, Engineering Failure Analysis. 28, 149–159, 2013. 113  Ling Yun, Z. G. Chen, P. Guempl, M. Kaesser M, Influence of Water Pollution on MIC of Stainless Steel 304L, 钢铁研究学报:英文版. 9, 45-48, 2002. E.M. Zahrani, A. Saatchi, A. Alfantazi, Pitting of 316L stainless steel in flare piping of a petrochemical plant, Engineering Failure Analysis. 17, 810–817, 2010. K. Zhao, Investigation of microbiologically influenced corrosion (MIC) and biocide treatment in anaerobic salt water and development of a mechanistic MIC model, 2008. Zhao, Kaili. Investigation of Microbiologically Influenced Corrosion (MIC) and Biocide Treatment in Anaerobic Salt Water and Development of A Mechanistic MIC Model, PhD Dissertation. Ohio University, 2008.  K. Zhao, T. Gu, I. Cruz, A. Kopliku, Laboratory investigation of MIC in hydrotesting using seawater, NACE Corrosion Conference, 2010.   114  APPENDIX   Tensile Testing Results for the first sample   115   Tensile testing result for the second sample   116   More samples from the piping network experienced the same failure morphology. 117   Another sample from the failed piping network experienced the same failure morphology  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0366239/manifest

Comment

Related Items