Underdeposit corrosion (UDC), a general term that refers to localized corrosion that develops beneath or around deposits on a metal surface, is a phenomenon that leads to pipeline failures and is one of the most damaging forms of corrosion for oil and gas pipelines. It can occur in subsea water injection pipelines, well-fluid pipelines, and large-diameter transmission lines.
According to Pavan K. Shukla with Savannah River National Laboratory (Aiken, South Carolina, USA) and Sandeep Naraian, an independent corrosion consultant (Navi Mumbai, Maharashtra, India), UDC is typically localized and very aggressive and can lead to deep penetrations (pits) in a metal surface that are accompanied by general corrosion in the surrounding areas. In some cases of UDC, pitting extends throughout the entire metal surface, which gives it an irregular or very rough surface profile. In other instances, pits are concentrated in specific areas, which leaves the majority or the metal surface in like-new condition. Microbiologically influenced corrosion (MIC) is also associated with UDC and may be linked to the development of internal deposits within the piping system. The presence of bacteria populations can produce an acidic environment that helps to corrode metal piping at highly accelerated rates—exceeding 75 mpy (1.9 mm/year).
Shukla and Naraian reported on the case of an American Petroleum Institute (API) 5L Grade X-60 carbon steel injection water pipeline that failed four years after commissioning due to UDC. The 9.75-in (248-mm) internal diameter pipeline was commissioned in early 2008 with a design life of 25-plus years. The maximum operating pressure and temperature were 105 kg/cm2 (1,494 psi) and 35 °C (95 °F) respectively. According to Shukla and Naraian, seawater is the main water source for injection in the reservoir for enhanced oil recovery in offshore systems. In a typical water injection process system, water is lifted from a depth of 30 m (98 ft) from the sea level and processed prior to injection to ensure its compatibility with produced water (i.e., it does not form precipitates).
To investigate the cause of failure, a detailed study was carried out that examined pipeline materials; injection water parameters, such as chloride content, suspended solids, turbidity, microbial counts of sulfate-reducing bacteria (SRB), dissolved oxygen (DO) levels, iron contents, and chemical additions (e.g., ammonium bisulfite [ammonium hydrogen sulfate] [(NH4)HSO4], an oxygen scavenger); pigging operations; and use of a water injection corrosion inhibitor.
Seawater analysis indicated it contained 20,590 ppm of chloride ions, which are known to promote pitting and other forms of localized corrosion because they break down protective films on steel surfaces. Shukla and Naraian note that chloride content in the range of 10,000 to 100,000 ppm will increase the corrosion rate. The effectiveness of chemical corrosion inhibitors also can be reduced in the presence of chloride, they add. If the level of chloride ions in a water injection process system is high (>30,000 ppm), inhibitors and inhibition procedures should be carefully selected and performed.
When bacteria are in close contact with a metal surface and form a biofilm, the microbial activities can affect the corrosion reactions as well. There are several groups/types of micro-organisms that are responsible for MIC on the internal pipeline surface, including iron-oxidizing bacteria and sulfur-oxidizing bacteria as well as SRB. Shukla and Naraian explain that SRB require an environment without oxygen to function most efficiently, and their growth is facilitated in a temperature range of 20 to 40 °C (68 to 104 °F), with maximum growth at 35 °C (95 °F). The bacteria form colonies in low-velocity/stagnant flow areas, particularly along the bottom of the pipeline.
In this water injection pipeline system, the SRB count at the pipeline entry was mostly negative; however, at the well-head end, the count was observed to be in the range of 103 CFU/mL, which is elevated. One possible reason for the higher SRB count at the well-head end is the sudden change from an aerobic to anaerobic environment downstream after deaeration (in the de-oxygenation tower). This anaerobic condition is more favorable for certain bacteria growth, including SRB.
Loosely bound black and gray corrosion products, as well as scales, were found on the internal surface of the pipeline but were less prominent in areas near and around where the failure occurred. The deposits were analyzed by wet chemical methods for the presence of various anions such as chloride, carbonate, sulfide, and oxides; and the results indicated mainly the presence of carbonates, sulfide, iron oxides, and traces of chloride.
An examination of the failed pipeline revealed accumulation of deposits on the internal surface. According to Shukla and Naraian, the fluid velocity in the pipeline was in the range of 0.53 to 0.93 m/sec (2 to 3 ft/sec) (below the minimum required velocity of 1 m/sec, or 3 ft/sec) and not adequate to maintain the solids in suspension. This low velocity led to the internal accumulation of deposits along the bottom of the pipeline. High chloride content (20,590 ppm) in the injection water facilitated pitting corrosion by locally breaking down the protective film.
Visual inspection showed loosely adhered reddish-brown deposits and scales on the inner surface of the pipe. Wet chemistry and energy-dispersive x-ray spectroscopy (EDX) analysis of the deposits confirmed the presence of iron oxides and iron sulfide. Both iron oxide and iron sulfide accumulated at the 6 o’clock position along the pipeline’s internal circumference due to low fluid velocities.
Continuous measurements of DO determined its presence in a higher concentration than the minimum specified limit of 20 ppb. The oxygen beneath the deposits may have been consumed by a primary corrosion reaction or utilized by aerobic bacteria. Then, in the absence of oxygen, SRB activity flourished in the system and settled in sessile concentrations of 10 to 103 CFU/m on the internal metal pipe surface. Shukla and Naraian note the SRB likely grew with time, developing colonies and biofilms under the deposits that provided an excellent habitat for these bacteria and shielded them from effective treatment from biocides. The SRB activity resulted in production of hydrogen sulfide (H2S), which reacted with pipe metal and formed iron sulfide, which was confirmed in the deposited corrosion product analysis.
According to Shukla and Naraian, all these factors—the presence of sulfide, iron oxide, chloride, and calcium scale; MIC due to SRB; and various corrosion scales that formed on the inner pipeline surface—contributed to the UDC of the water injection pipeline. This UDC resulted in channeling or grooving corrosion along the bottom of the pipeline, which weakened the pipeline section. Without remedial actions the condition became worse and led to the development of cracks and a hole that caused the failure.
More details on the failure analysis can be found in CORROSION 2017 paper no. 8973, “Under-Deposit Corrosion in a Sub-Sea Water Injection Pipeline—A Case Study” by P.K. Shukla and S. Naraian, which is available from the NACE International web site at https://store.nace.org.
Contact Pavan K. Shukla, Savannah River National Laboratory—email: firstname.lastname@example.org.