Accelerated sulfate reducing bacteria corrosion of X80 pipeline steel welded joints under organic carbon source starvation

Long-distance oil and gas pipelines operate in complex, often anaerobic or hypoxic soils where microbiologically influenced corrosion (MIC) is a major threat. SRB-driven MIC is prevalent because SRB can thrive under such conditions. In nutrient-rich environments SRB reduce sulfate by oxidizing lactate; however, under mature biofilms organic carbon sources become scarce, and SRB shift to obtaining electrons from extracellular iron while reducing intracellular sulfate. This extracellular electron transfer (EET) mechanism makes the iron matrix an electron donor and drives corrosion. Welded joints comprise zones (base metal, BM; heat affected zone, HAZ; weld zone, WZ) with distinct microstructures and electrochemical activities that influence bacterial adhesion and electron transfer, leading to selective corrosion. Prior studies indicate SRB become more aggressive under organic carbon source starvation and that welded joints exhibit zone-selective MIC. This study asks two key questions: (1) When carbon-starved SRB encounter a welded joint, do they use all zones or specific zone(s) as electron donor(s)? (2) How does the corrosion behavior differ between the selected electron-donor zone(s) and other zones of the welded joint under organic carbon source starvation? The work aims to clarify SRB behavior and mechanisms of corrosion in X80 pipeline steel welded joints under carbon-starved conditions.

The paper summarizes prior findings that SRB under organic carbon source starvation accelerate corrosion of carbon steels, increasing pit depths and corrosion rates (e.g., C1018 and L245 steels). Starving SRB utilize the iron matrix as an electron donor to maintain metabolism via EET. Studies comparing standard versus starved media reported higher weight loss at reduced carbon source levels and greater aggressiveness of pre-established SRB biofilms under starvation. For welded joints, microstructural changes after welding affect bacterial attachment and result in selective corrosion: WZs with many grains and grain boundaries may attract more bacteria than BM, and SRB corrosion of X80 welded joints tends to be most severe in WZ and BM, with HAZ less affected. These works motivate examining how starvation modulates selectivity across BM, HAZ, and WZ and the relative contributions to uniform versus localized corrosion.

Materials and sample preparation:

• Material: Spirally welded joint of X80 pipeline steel. Composition (wt.%): C 0.0466, Si 0.204, Mn 7.154, P 0.0082, S 0.0009, Ni 0.206, Cr 0.235, Cu 0.174, Nb 0.524, V 0.0022, Ti 0.0142, Mo 0.125, Al 0.0265, B 0.0004, Fe balance.
• Zones prepared by wire cutting: Base metal (BM), heat affected zone (HAZ), and weld zone (WZ). Due to small WZ/HAZ areas, weight loss and sessile counts were performed only for BM.
• Surface preparation: Polished 280#→1200# SiC paper, rinsed, ethanol cleaned, air dried. UV sterilized in a vacuum glove box.

Microorganism and media:

• SRB strains consistent with prior work; cultivated in ATCC 1249 medium (3rd generation used).
• Modified ATCC 1249 media prepared with three carbon source levels by adjusting trisodium citrate and sodium lactate to 0%, 1%, and 100% of full strength; yeast extract present in all media providing small amounts of organic carbon and growth factors.
• Media sterilized (121 °C, 30 min) and deoxygenated with high-purity nitrogen for 2 h.

Experimental design:

• Pre-incubation to establish biofilm: All samples were pre-immersed 3 days in 1 L full carbon source medium containing SRB to form mature biofilms. Planktonic cells on surfaces were removed by sterile, deoxygenated PBS rinse.
• Starvation exposure: Biofilm-bearing samples were transferred into anaerobic flasks with 1 L media at 0%, 1%, or 100% carbon source and sealed. Immersion duration: 30 days under nitrogen.

Microbiological measurements:

• Planktonic SRB counts measured periodically using KBC test flasks designed for MPN (GB/T 14643.5-2009); reported as cell ml−1.
• Sessile SRB counts (BM only): After 30 days, samples were shaken in 100 mL deoxygenated PBS for 2 h; MPN measured and reported as cell cm−2.

Corrosion rate and morphology:

• Weight loss (BM only): Initial and final masses recorded (0.0001 g accuracy). Corrosion products removed with inhibitor-containing HCl solution. Corrosion rate Vcorr (mm y−1) computed by Vcorr = K × Δm / (ρ × S × t), with K = 87,600; ρ = 7.85 g cm−3; S = area; t = immersion time (h).
• Surface/product characterization: After 30 days and biofilm fixation (5 wt.% glutaraldehyde, 5 h, ethanol dehydration), corrosion products analyzed by XPS (ESCALAB250XI; C 1s at 284.6 eV reference; peak fitting via XPSPEAK4.1), SEM (Zeiss EVO), and EDS (Oxford X-Max). Corrosion morphologies observed by SEM after product removal. 3D ultra-depth microscopy (VHX-2000) used to reconstruct pit topology and measure pit depths.

Acceleration and selectivity metrics:

• Uniform corrosion acceleration factor: K_CR = CR / CR0, where CR is BM corrosion rate under starvation (0% or 1%), CR0 is under 100% carbon.
• Pitting acceleration factor: K_pit = l / l0, where l is average pit depth under starvation and l0 under 100% carbon.
• Selectivity factor: K_select = I_i / I_HAZ, where I_i is average pit depth in BM or WZ and I_HAZ is average pit depth in HAZ within the same medium. By definition K_select(HAZ) = 1 in each medium.

• Cell dynamics:

  • Planktonic SRB in all media exhibited limited-environment growth curves, peaking on day 3. Peak counts in 100% carbon medium were 1–2 orders of magnitude higher than in 0% and 1% media. Counts declined to ~10^4 cell ml−1 by day 15 and decreased slowly thereafter.
  • Sessile SRB on BM after 30 days were highest at 1% carbon, then 0%, and lowest at 100%. Sessile counts greatly exceeded planktonic counts at experiment end. No direct correlation between planktonic and sessile numbers.

• Macroscopic and microscopic observations:

  • 0% and 1% media: Numerous microbial tubercles on BM; dense corrosion spots aligned with tubercle locations after product removal.
  • 100% medium: BM surface relatively flat; no obvious corrosion spots.
  • Corrosion products distribution: Uneven and flake-like (0%); clustered (1%); relatively uniform with few prominences (100%).

• Corrosion rates (BM, 30 days): Highest at 1% carbon, followed by 0%, and lowest at 100%, matching sessile SRB abundance. Organic carbon starvation increased SRB aggressiveness.

• Product chemistry (XPS/EDS):

  • 100% carbon: Corrosion products mainly FeOOH; higher sulfur content detected relative to starvation conditions by EDS.
  • Starvation (0% and 1%): Broader iron oxide species including FeOOH, Fe3O4, and Fe2O3; Fe and O contents increased in products. Sulfur species including FeS and FeS2 (typical SRB metabolites) and FePS3 detected in all media. Various organic nitrogen-containing functionalities consistent with biofilm components were present.
  • SRB cell morphology in product films: More numerous and often slender/stubby under starvation; mostly stubby in 100% carbon—indicating relatively higher metabolic activity in starvation.

• Localized corrosion and selectivity:

  • Across all media, localized pitting dominated; BM and WZ consistently had larger pit diameters/densities than HAZ. In 100% carbon medium, HAZ exhibited almost no pits.
  • Acceleration factors (BM): K_CR = 1.15 (0%); 1.26 (1%). K_pit = 1.75 (0%); 3.33 (1%). Starvation accelerated both uniform and localized corrosion, with a much stronger effect on pitting (K_pit >> K_CR), especially at 1% carbon.
  • Selectivity factor K_select: For BM and WZ, K_select > 1 in all media (HAZ = 1 by definition). K_select(BM) and K_select(WZ) were larger under starvation than at 100%, and largest at 1% carbon. Order of susceptibility remained BM > WZ > HAZ in all media.

• Biofilm pre-incubation: Mature biofilms formed after 3 days in full carbon source were similar across zones; by 7–14 days, BM and WZ developed more microbial tubercles than HAZ, correlating with localized corrosion beneath tubercles.

The study demonstrates that organic carbon source starvation increases SRB reliance on extracellular electron transfer from the iron matrix, thereby accelerating corrosion of X80 welded joints. This acceleration is markedly stronger for localized pitting than for uniform corrosion, as evidenced by K_pit values greatly exceeding K_CR under starvation. The effect peaks at 1% carbon source, where SRB can obtain electrons from both limited organic carbon and the iron matrix, promoting more sessile cell attachment, more microbial tubercles, and more severe pits. Selectivity arises from microstructural and electrochemical heterogeneity among welded zones. BM and WZ, featuring microstructures with higher intrinsic energy, more grain boundaries, and lower electron work functions, provide more favorable sites for bacterial adhesion and electron uptake than HAZ. Consequently, BM and WZ act as preferential electron donors and experience deeper and denser pitting. Organic carbon starvation amplifies this intrinsic selectivity (higher K_select for BM and WZ) without changing the susceptibility order (BM > WZ > HAZ). Surface chemistry corroborates active SRB metabolism and EET under starvation (mixed iron oxides and SRB-associated sulfides) and supports the link between sessile SRB abundance and corrosion severity. Overall, the findings answer the research questions by showing that carbon-starved SRB preferentially utilize specific zones (BM and WZ) as electron donors, leading to more severe localized corrosion in these zones compared with HAZ, with the strongest effect at 1% carbon source.

Organic carbon source starvation makes SRB more aggressive toward X80 pipeline steel welded joints. Starvation increases both uniform and especially localized corrosion, with the most severe effects at 1% carbon source where sessile SRB are most abundant. Corrosion is strongly zone-selective: BM and WZ are preferentially attacked compared to HAZ due to microstructural factors that favor electron transfer and bacterial adhesion. Starvation further enhances this selectivity (higher K_select) but does not change the susceptibility order (BM > WZ > HAZ). These insights improve understanding of SRB EET-driven MIC in welded joints and highlight the need for further investigation into how microstructure and electron work function govern SRB adhesion and electron uptake in realistic service environments.

• Weight loss measurements and sessile SRB counts were performed only on BM samples due to the small sizes of WZ and HAZ zones.
• The 0% carbon source medium still contained a small amount of organic carbon from yeast extract.
• Pre-established biofilms were formed under a specific pre-incubation condition (3 days in full carbon source), which may influence subsequent corrosion behavior.
• Experiments were conducted on a single welded joint material (X80) and with a specific SRB strain in controlled laboratory media and anaerobic conditions.