Authors: Daniel Sandana
Ensuring Safe and Reliable Transport of Anthropogenic Dense CO₂: Advancing Integrity Management and Energy Security through Data-Driven Insights
In a nutshell:
The definition of energy security is changing. Reliable oil and gas supply is critical, but energy systems must also support the transition to lower-carbon solutions without compromising safety or continuity of supply. As CCUS becomes central to this transition, the transport of anthropogenic CO2 introduces new pipeline integrity risks that conventional approaches do not fully address. Ensuring safe operation will require a step change in integrity management through next-generation in-line inspection solutions and AI-supported data analyses, including approaches such as ROSEN's MFL Data Fusion.
The CCUS industry has a long history. The first large scale CO2 pipeline, known as the Canyon Reef Carrier, a 16” – 222.4 km pipeline, has been in operation since 1972. Much of the remainder of the current United States (US) CO2 transportation infrastructure was built between the 1980s and 1990s, and stretches to approximately 7000 kilometers. Decades of experience with dense-phase CO2 pipelines in North America present a statistically positive and reassuring safety record. There have been no fatalities associated with these pipelines. In addition, fewer than 5% of reported incidents have released more CO2 than a transatlantic flight between London and San Francisco, and fewer than 30% have released more CO2 than a large-scale event such as a Taylor Swift concert. While the latter are recurring accepted public events, CO2 pipeline incidents remain rare, with fewer than 120 accidents recorded since 1991.
The next generation of CO₂ pipelines will operate under different conditions
It is then only natural to question the actual concerns of operating safely dense CO2 pipelines going forward. A key aspect of the existing operational US CO2 pipelines is that they transport CO2 captured from natural gas production, ammonia and ethanol productions. But, as we shift to the capture of CO2 from industrial sources with a wider variety of impurities, new challenges emerge. The core of the next generation of CO2 pipeline projects will aim at transporting man-made CO2. which will carry a broader number of impurities. A number of these impurities have received particular attention (e.g. SOx, NOx, H2S, O2 – but may not be limited to) as they can have significant implications for the realization of pipeline internal corrosion, and how this is safely managed during design and operation.
Understanding corrosion behavior in CO2 pipelines
The management of internal corrosion in pipelines transporting anthropogenic CO2 is a critical topic. Much research has been carried out to establish safe and practical specifications (composition limits) to address the integrity challenges (e.g., acid drop-out) posed by contaminants such as SOx and NOx. However, gaps remain and findings are bound to experimental challenges, artifacts and limited test conditions (e.g., simple mixtures, pressure, temperature), which may lead to unsafe or overly conservative interpretations. The uncertainty is significant as there is still no operational pipeline transporting anthropogenic CO2 containing this class of contaminants. The key question is then how we carry the residual risk into operations and how to balance ”leap of faith” against risk.
A key part of the equation lies in the use of ILI to demonstrate that the composition limits applied are safe and ensure operational integrity over the pipeline life cycle. Even then, the occurrence of internal corrosion in CO2 pipelines presents unique challenges, that drive towards enhanced detection and sizing accuracy. In addition the ability to characterize corrosion profiles becomes vital.
As CO₂ transport becomes critical to the energy system, integrity engineering must evolve from inspection to insight. Understanding complex corrosion mechanisms is no longer just a technical challenge – it is becoming central to how we ensure energy security in a decarbonizing world.
Complex corrosion morphologies
The physical and chemical processes at play, compounded by the presence of impurities e.g. SOx, NOx, H2S, O2, mean that damage primarily manifests as discrete pits (Figure 1: (a), (b)). Over time, pits grow laterally to evolve into clusters and progress into more generalized, uniform corrosion. These dynamics lead to the formation of complex corrosion profiles, with abrupt and irregular reliefs in axial and circumferential directions, with deeper points (‘pits-in pits’) (Figure 1: (c)).
Many standard ILI techniques and their associated data analysis models are not designed to resolve the intricate details of such complex corrosion, potentially missing key features and not fully characterizing the corroded area. In addition, dense phase CO2 is typically transported at pressures between 120 to 200 bar, which is generally associated with high operational stresses e.g. >60% SMYS; this increases the need for detecting and sizing accurately small anomalies to ensure safety.
Aggressive growth rates
The presence of an aqueous phase at the bottom of a CO2 pipeline can lead to unmanageable aggressive corrosion rates; rates well in excess of 10-20 mm/year have been quoted. In reality, the corrosion rates commonly quoted reflect extreme scenarios associated with uncontrolled continuous upsets, i.e. uninterrupted replenishment (drop-out) of aqueous phases and corrodents (e.g. strong acids). These circumstances are unrealistic in practice, operationally not manageable and must be avoided through strict adherence to targeted CO2 quality specifications.
That said, temporary operational disruptions and off-spec CO2 remain a realistic and distinct possibility, particularly in clustered system configurations with multiple feeders of varying characteristics, which increases the likelihood of transient abnormal events. Some (limited) data points indicate that average corrosion rates (CRs) associated with accumulated stagnant water pools in dense CO2 conditions could remain substantial, for example, 0.5- 1 mm/yr, although these are not strict lower or upper bounds. The modeling of CRs in transient upset scenarios and unreplenished liquid hold-ups is a key industry gap.
The aggressiveness of the corrosion significantly reduces the tolerance to upset conditions, and puts further emphasis on early detection of developing localized corrosion cells in order to address uncertainty in integrity management. As the margin of safety can be reduced due to the high operational stress and upper end corrosion rates, more accurate sizing in combination with less conservative Fitness-for-Service (FFS) assessments is also necessary to optimize inspection planning and avoid over conservatism and unnecessary remedial actions (digs and repairs).
The limitation of traditional in-line inspection approaches
Given the challenges of internal corrosion in CO2 transportation the impetus is there for ILI with enhanced detection capability, ability to characterize complex shapes and feature profiles, and deliver a high level of sizing accuracy, Ultrasound Testing (UT) would offer, in theory, a more pertinent technology. However, although the dense CO2 phase reacts like a liquid with regard to some of its physical properties, the application of liquid coupled UT is not possible with the required accuracy. This is primarily due to the high level of variation in the density, sound velocity and impedance that occurs in dense phase or supercritical fluids with any change in temperature and pressure, which is unavoidable in a dynamic pipeline environment, and the passage of a pig. These variations affect the behavior of the UT beam, leading to unpredictable and unreliable inspection results. For dense phase applications, the preferred method of inspection is then MFL-based, but it comes with its own limitations that reduce confidence in integrity management of CO2 pipelines.
Magnetic Flux Leakage (MFL) technology is widely used to detect metal loss in pipelines, with axial (MFL-A) and circumferential (MFL-C) methods each suited to different defect orientations. However, both rely on a single magnetic field direction, limiting their ability to accurately detect complex corrosion shapes. To improve detection, both methods are often used together, but their results are analyzed separately, which can be inefficient, subjective, and sometimes conflicting. Conventional reporting also simplifies corrosion into rectangular “boxes,” losing detailed shape information and leading to conservative assessments that do not support optimal decision-making for integrity engineers responsible for anthropogenic CO2 pipelines.
To overcome the limitations with analysis and reporting of combined MFL data, the development of methods or processes that can fully exploit the complementary data available from both MFL-A and MFL-C is beneficial to overcome areas of subjectivity, reduce uncertainty, and aim for more reliable and pragmatic integrity management of CO2 pipelines. One path is the use of AI-enabled analysis such as MFL Data Fusion.
MFL Data Fusion enhances integrity management of CO2 pipelines
The MFL Data Fusion process is carried out using the MFL signal data to overcome the subjectivity in the combined inspection report. The fused data can deliver a level of insight that surpasses the capabilities of either technology alone. The raw measured axial and circumferential MFL signals from each respective inspection technology are taken as inputs. Pre-processing is first carried out on each data set to eliminate variations in the signals and remove signal noise. Before the data can be fused, it is essential to carry out detailed alignment of the two data sets to achieve pixel level matching. This provides a two channel image which is used to achieve accurate alignment using a matching algorithm.
The aligned MFL signals are then input into the machine learning Data Fusion model. The fusion model is a U-Net architecture neural network which has been pre-trained on simulated MFL data and laser scan data. Further validation and refinement of the MFL fusion model can be done based on field data e.g. laser scan or automated ultrasound testing. A single output is produced which is a high resolution 3D metal loss depth map. This can be used to generate detailed anomaly profiles along the pipeline length (Figure 2).
For the integrity management of anthropogenic CO2 pipelines, the benefit of such approach is multiple:
- First, it provides a step change in the ability to detect and characterize complex corrosion morphologies. In environments where damage often initiates as highly localized pits and evolves into irregular “pit-in-pit”, with MFL Data Fusion, operators gain access to a more representative and high-resolution reconstruction of metal loss, enabling a more reliable understanding of defect morphology and severity. This is particularly critical in the presence of localised corrosion cells, where early-stage damage may otherwise remain undetected until it reaches a critical size.
- Second, the generation of detailed 3D corrosion profiles fundamentally improves the quality of integrity assessments. Rather than relying on simplified geometrical assumptions, operators can base Fitness-for-Service evaluations on more realistic representations of corrosion features. This reduces unnecessary conservatism while maintaining safety margins, a key requirement in dense-phase CO₂ pipelines operating at high stress levels and with limited tolerance to upset conditions. This ultimately improves the operator’s agility to manage the tolerance of upset conditions more pragmatically, thereby enabling the development of more rational / practical integrity management (e.g. cleaning pigging response, tolerated frequency of upsets, in-line inspection frequency) without compromising the integrity strategy.
Why this matters for energy security
As CO2 pipelines become part of the energy backbone, energy security will increasingly depend on how well their integrity risks are understood and managed. AI-supported data analyses does not just improve measurement accuracy – it fundamentally changes how operators assess risk, prioritize actions, and plan mitigation. By enabling deeper insight into complex degradation mechanisms and reducing uncertainty in integrity assessments, these approaches support more informed, timely, and proportionate decision-making. In doing so, it enables safer, more resilient CO2 transport systems – ensuring that decarbonization infrastructure operates reliably as part of the wider energy system.
References
- Yoon-Seok Choi, Fernando Farelas, Srdjan Nešic, Alvaro Augusto O. Magalhães,* and Cynthia de Azevedo Andrade, “Corrosion Behavior of Deep Water Oil Production Tubing Material Under Supercritical CO2 Environment: Part 1—Effect of Pressure and Temperature”, Corrosion, 2014
- Fareles, Choi, Nesic, “Effects of CO2 Phase Change, SO2 Content and Flow on the Corrosion of CO2 Transmission Pipeline Steel”, Corrosion, 2012
- Daniel Sandana, Angus Patterson, Kevin Siggers, Ensuring safe and reliable transport of anthropogenic dense CO2: advancing integrity management with MFL Data Fusion, IPC 2026-185999, Calgary, 2026
- Xiang, Peng, Kevin, Siggers, Mark, Wright, Johannes, Palmer, “Data Fusion of Complementary Axial and Circumferential Magnetic Flux Leakage Inline Inspections and Effects on Safe Remaining Life”, IPC 2024, Calgary, 2024
Daniel Sandana
Principal Materials and Corrosion Engineer
Daniel Sandana, Principal Materials and Corrosion Engineer at ROSEN, holds an MSc in Materials Science and Engineering from ESIREM in France and a PhD in Metallurgy/Corrosion from Newcastle University in the UK. He is also a European Chartered Engineer (Eur Ing).
Daniel has over 20 years of experience in asset integrity management in the global oil and gas upstream and transmission sectors. Since 2009, he has been involved in CO2 transportation research, contributing to early European carbon capture and storage initiatives.
Daniel is currently involved in industry initiatives that support operators in transitioning to decarbonized energy systems. He has authored over 50 peer-reviewed technical papers and regularly provides industry training on the safe transportation of CO2 and H2.