Electrical isolation in pipelines or piping occurs when one of the following takes place:
• A dissimilar metal joint is in the presence of an electrically conductive fluid.
• A transition occurs from an underground to an aboveground section of a cathodically protected pipeline when the aboveground section is grounded at some point.
• The amount of cathodic protection (CP) current required is limited to protect the structure and achieve the right level of polarization.
A description of the different devices that can be used to achieve electrical isolation can be found elsewhere.1 The selection of the specific isolating device depends on the mechanical forces exerted on the line, pressure constraints, and maintainability, among other considerations.
Monolithic isolation joints (MIJs) are one type of isolating device. Basically, an MIJ consists of a pair of short pipe lengths; one is extended at its periphery by a barrel that overlaps the other. The two sections are isolated from each other by nonmetallic materials that are restrained and maintained in position by large compressive forces locked by welding, wedging, or swaging and are pressure sealed. There are advantages to these devices, such as they do not contain threaded components; they are factory assembled, reducing the risks associated with installation; and they are designed and produced according to pipeline codes and specifications.
MIJs are complex items as they congregate a series of materials with different properties, notably steel, elastomers, glassreinforced epoxy (GRE), and coatings. The properties of all these materials must be considered when producing and assembling an MIJ, and matching their properties with the operation envelope is paramount to ensure a good performance. A recent paper reviewed the requirements that must be addressed to ensure the proper performance of an MIJ.2 These included the limitations associated with isolating devices, which are essentially related to a high degree of quality assurance/quality control (QA/QC) required during construction, assembly, and installation, given the different materials involved.
Loss of containment at MIJs were related to failure of the compression set of the O-ring, cracked energized springs, excessive hard welds, low QA/QC performance during MIJ fabrication, and design flaws when the electrical isolation sheet is used simultaneously as an isolation component and as a critical sealing component.3 Other possible causes were attributed to sealant failure due to application of the isolating joint outside of its design operating temperature.4
Failures associated with a properly designed MIJ can also result from current bridging due to internal coating damage, such as the result of scraping pig operations or coating abrasion when sand production occurs with the produced fluid.
This article describes the investigation of a leak that occurred on a buried threelayer polypropylene-coated oil transfer line just seven months after commissioning of the CP system. The leak was detected at the 6 o’clock position, ~1 m downstream of the isolating joint, at the receiving end of an oil transfer line.
The 24-in monolithic isolating joint was installed above ground, on a horizontal plane, and was made of cold drawn seamless (CDS) carbon steel (CS) (ASTM A1065) with insulating spacers made of GRE (grades G10-G11).6 The sealing consisted of an elastomeric material suitable to meet the specified design conditions.
Spark arrestors were installed across both isolating joints, located at the ends of the pipeline.
The pipe section corresponding to the nonprotected side of the MIJ was cut longitudinally to allow the observation of the internal surface (Figures 1 and 2).
The visual observation of the internal surface revealed:
• The internal coating was in good condition and extended up to 90 cm from each side of the joint.
• A section with several pits distributed on an arc of circumference between the 4 to 5 and 7 to 8 o’clock positions on a strip delimited by the end of the internal coating and the circular weld connecting the isolating joint to the CDS pipe side (Figure 3), where the trespassing pit was located at the 6 o’clock position, 30 mm downstream of the coating. The pit distribution extended up to ~1 m of the coated section, and the number as well as the depth of the pits faded as the distance increased, converging to the 6 o’clock position (Figures 1 and 2). All pits presented a cylindrical shape, more evident as their number decreased.
• Corrosion damage occurred at the steel/dielectric material interface, on the CDS (unprotected) side of the MIJ between the 4 and 8 o’clock positions, with through corrosion of the pipe thickness (Figure 3). No leak has been observed at this location, which was attributed to the geometry of the MIJ.
Laboratorial testing of metallic samples collected from both pipe spools revealed chemical and mechanical properties in line with the requirements of ASTM D709.7 Metallographic examination of the weld joints revealed microstructures coherent with an adequate welding technique. Hardness test results for the different zones of the weld were found below 245 HV 10, in line with the sour service requirement.8
Review of the swing test results performed during commissioning did not highlight any lack of isolation.
Discussion
Engineering design considered a produced multiphase fluid with water content of 29%. A water resistivity of 5 Ω·cm was estimated based on a chloride content of 13.1% measured in the produced water.9 As per engineering best practices, the electrical isolation required an isolating spool.
A literature survey revealed there are several methodologies to calculate the required length of an internally coated pipe section to prevent bridging of current from the unprotected to the protected side of the MIJ. The industry-accepted best practice relies on an empirical correlation between the electrolyte resistivity and the pipe diameter10 to determine the required spool length. This practice was compared with alternative models that rely on either an electrical attenuation model applied to the internal side of the pipe11 or a model based on an electrochemical model considering activation polarization and large overpotential on the steel surface.
Applying the engineering best practice, the estimated spool length was 73 m, the attenuation-based model provided a required internal coating length of the isolating joint of 8.7 m, while the electrochemical-based model required a coated length of 38 m. The total applied internally coated length of the isolating joint was 1.8 m (0.90 m to each side). As the comparison between this value and the previous calculations has shown, regardless of the model used, the MIJ was not adequately designed, and current was expected to bridge across the MIJ, producing the corrosion observed downstream of the coated section of the MIJ (unprotected side). The corrosion observed immediately downstream of the internally coated section at the unprotected side of the MIJ was caused by current flowing between the unprotected and protected sides through the internal conductive media, which is a well-known phenomenon.
The step resulting from the weld root penetration associated to a low-flow condition of the fluid contributed to a semi-circumferential distribution of the pitting upstream of the field weld joint and the confluence of pitting distribution downstream of the weld.
A significant amount of corrosion was also observed at the steel/isolating material interface on the unprotected side of the isolating joint, indicating that the gap between the nonmetallic material and the face of the steel was compromised, leading to loss of isolation in this area, further reducing the isolation length. The break of isolation at the steel/nonmetallic interface by electrolyte ingress may have resulted from:
• Bending stresses caused by improper installation of the isolating joint. In fact, inspection of the isolating joint while still assembled on the pipeline revealed that it was not properly installed. As per the manufacturer’s recommendations, it should have been installed in a pipe support to avoid bending momenta. Lack of support introduced a momentum caused by gravity, inducing stresses in the O-ring and/or the isolating elements, which was conducive to electrical contact at this location.
• Tensile or bending stresses applied on the monolithic isolating joint caused by buckling of the line.
• Deficiency/absence of the coating at the metal-to-polymer interface or sealant coating applied on the internal surface for proper sealing of the isolating material.
• Damage to the O-ring or the isolating element due to poor workmanship, notably excessive or uneven radial compression or excessive heating of the nonmetallic elements during welding operations.
• Poor design or workmanship of the nonmetallic resin. In the present case, the glass transition temperature of the epoxy was very close to its design temperature, eventually contributing to the failure by compromising the dielectric properties of the joint and promoting an electric path across the MIJ.
The damage observed at this interface has not yet been reported in the literature, although it has a direct impact on the performance of the isolating joint by reducing the isolated section by one half of the isolated path (as it is equally internally coated on both sides). Loss of isolation at the isolating material interface probably resulted from bending stresses or buckling of the pipeline, which caused a reduction of the already under-designed coated section of the pipe, enhancing the amount of CP current bridging across the isolation, and therefore, a high corrosion rate of the steel.
The observation of the leak in the pipe, and not at the isolating joint itself, is justified by the geometry and overall thickness of the steel at the MIJ. Based on geometrical considerations, the distribution of the corrosion at the 4 and 8 o’clock positions on the unprotected (anodic) side of the MIJ, next to the isolating material, suggest a level of water inside the pipe, corresponding to ~20% of the pipe section. The corrosion features observed at these positions can be justified by the fact that part of the current carried through the pipe by the electrolyte discharges at this location, increasing the local anodic current density.
If the current leakage would have been captured by early CP data gathering; for example, by testing the isolating joint with the four-point method, its effects on the unprotected side of the pipe immediately downstream of the internally coated section could have been detected in-service by manual ultrasonic thickness (UT) scanning of the lower half of the pipe. The corrosion observed at the MIJ itself is much harder to detect, requiring a skilled UT operator.
Conclusions
The failure resulted from corrosion due to current leakage across the protected and unprotected sides of the monolithic isolating joint. This was caused by a deficient design of the MIJ and the electrical properties of the produced fluids. Stresses applied on the MIJ and an inadequate material specification for the nonmetallic resin used in the isolating material were contributing factors.
The design calculation of the isolating spool length should be clarified, as existing models present rather different results. Engineering best practices should consider issues such as:
• Lack of space to implement the required spool length.
• Limiting the use of internally coated CS spools longer than a pipe length, as internal coatings of the weld areas represents a challenge for the current stage of technology.
• The use of GRE installed above ground is not allowed by some operating companies, presenting a further challenge for pressure rating and/or temperature.
References
1 NACE SP0286-2007, “Electrical Isolation of Cathodically Protected Pipelines” (Houston, TX: NACE International, 2007).
2 M. Monica, et al., “In Defense of the Monolithic Isolation Joint,” CORROSION 2017, paper no. 8993 (Houston, TX: NACE, 2017).
3 K. Doering, et al., “Monolithic Isolation Joints: A Possible Weak Link in Pipeline Integrity,” CORROSION 2014, paper no. 3989 (Houston, TX: NACE, 2014).
4 ADNOC Onshore, unpublished results, 2018.
5 ASTM A106, “Standard Specification for Seamless Carbon Steel Pipe for HighTemperature Service” (West Conshohocken, PA: ASTM International, 2018).
6 NEMA LI 1-1998 (R2011), “Industrial Laminated Thermosetting Products” (Rosslyn, VA: National Electrical Manufacturers Association, 2012).
7 ASTM D709, “Standard Specification for Laminated Thermosetting Materials” (West Conshohocken, PA: ASTM, 2017).
8 NACE MR0175/ISO 15156, “Petroleum and natural gas industries—Materials for use in H2S-containing environments in oil and gas production” (Houston, TX: NACE, 2015).
9 API RP 45, “Recommended Practice for Analysis of Oilfield Waters” (Washington, DC: American Petroleum Institute, 1998).
10 Shell DEP 31.40.30.39.
11 J.L. Brazy, “Dimensionnement d´une Manchette Isolante—Étude Theorique,” private communication, 1980.
12 W. von Baeckmann, W. Schwenk, W. Prinz, Handbook of Cathodic Corrosion Protection—Theory and Practice of Electrochemical Protection Processes, 3rd ed. (Houston, TX: Gulf Professional Publishing—Elsevier, 1997).