Background

Accurate assessment and prediction of corrosion of steels is one of the most significant tasks in current-day corrosion engineering. This is because steels are the primary materials of construction in just about every sphere of petroleum production, transmission and refining. While both metallurgical, manufacturing and economic considerations promote the utilization of steels, one inherent limitation of steels is their susceptibility to corrosion damage in environments that contain significant amounts of carbon dioxide, hydrogen sulfide and chlorides in combination with produced water. However, if data is developed to accurately model both the mechanisms and manifestations of corrosion damage, then it is possible to define conditions of safe utilization of steels as a function of operating environmental and metallurgical parameters. Further, a proper understanding of thermodynamics and kinetics of corrosion will allow us to develop meaningful guidelines for determining conditions for safe utilization of steels.

Development of corrosivity assessment and prediction criteria for steels requires determining two primary aspects of corrosion:

1. Mechanisms and underlying parameters that promote corrosion and accelerate metal loss, such as CO₂, H₂S, pH, chlorides, temperature, flow etc.
2. Parametric effects that mitigate corrosion severity such as scaling, hydro-carbon film persistence (oil wettability) and alloying effects.

It is necessary to characterize both individual parametric effects and parametric interactions in assessing corrosivity. Significant parameters in such an evaluation include,

• Carbon dioxide partial pressure
• Hydrogen sulfide partial pressure
• pH of aqueous phase
• Chloride concentration
• Temperature
• flow effects (velocity, shear stress)
• Hydrocarbon type/inhibition efficacy

CO₂-based corrosion has been one of the most active areas of investigation in industry and academia, with several predictive models for corrosion assessment currently available. However, most of these models suffer from significant drawbacks in that,

• They focus on a narrow range of parametric effects, for e.g., there has been little or no work on determining the important aspects of H₂S effects on corrosion and scaling of steels
• Some models focus on a narrow range of parametric effects, such as erosional effects or wall shear stress calculations or flow effects and often ignore the role of important chemical species (factors such as pH, H₂S etc.)
• Many models make simplifying assumptions not valid for field operations or use laboratory trends over a narrow range of parametric effects with little practical relevance.

InterCorr has developed an integrated corrosion prediction model and tool called Predict that addresses some of these concerns in that corrosion prediction in Predict is a function of all relevant environmental parameters. Further, the Predict system incorporates comprehensive corrosion prediction guidelines from several widely published models. A flow chart depicting the decision making steps used in the PREDICT system is shown in Figure 1.

Figure 1 -- Predict Flow Chart

However, there are still several aspects of corrosion prediction that are not addressed in the Predict system because of non-availability of data to characterize these critical effects. Such critical aspects include,

• Data to capture and understand critical issues in corrosion such as H₂S corrosion and scaling
• CO₂ / H₂S Equilibria and relative dominance of CO₂ or H₂S corrosion dynamics in production systems
• Role of hydrocarbon chemistry in promoting/inhibiting corrosion etc.
• Modes of corrosive attack on steels exposed to multi-phase CO₂ / H₂S environments as a function of environmental and metallurgical parameters
• Role of alloying, metallurgy and steel processing in corrosion mitigation
• Field data to calibrate and validate lab trends and modeling rules

The current testing and research effort is aimed at data generation to develop a firm basis for understanding and modeling the aforementioned critical issues. Also, the idea is to generate relevant data and facilitate creation of a computer-based tool to assess and predict corrosive severity with the above criteria as guiding principles. To that end, the research effort outlined herein is a logical extension of the significant corrosion modeling expertise available at InterCorr. InterCorr also has extensive experimental capabilities to monitor corrosion in the laboratory under simulated field conditions involving high pressure, high temperature and corrosive acid gases in combination with controlled flow conditions. This has provided the capability to examine:

• Effects of specific operating conditions that include corrosive chemical species and flow parameters
• Ability to dynamically model hydro-carbon film stability and determine efficacy of inhibition/chemical treatments
• The behavior of selected steels with varying composition, processing history and microstructure
• The interaction of metallurgical, environmental, and mechanical parameters so as to develop data and rules that can support application of steels under conditions normally reserved for CRAs