Strategy-A System Description

The Strategy-A system provides the user access to the material selection decisions and decision logic of a domain expert in evaluation of steels exposed to sour pipeline environments. The system also embodies information from other sources such as published literature on lab and plant experience related to pipeline service. The expert system is integrated with a Paradox database through the Strategy-A interface so as to facilitate storage of historical data.

Pipeline applications typically utilize welded construction of low to moderate strength steels. They also contain hydrocarbon media under pressure (which often contain H₂S and CO2 with low pH). Typical pipeline environments and the process of evaluating steels for service in such environments in Strategy are represented in terms of modular tasks described below:

1. Ranking of the pipeline environment in terms of severity from a stand point of hydrogen cracking to determine an Environmental Severity Factor (ESF) for HIC, SSC and SOHIC. Pipeline environments typically feature weak to moderate acidity solutions formed by a combination of H₂S and CO2 with water. The acid environment promotes corrosion and the H₂S enhances absorption of hydrogen into steel. Hydrogen flux in steels and associated cracking severity increases with decreasing pH and increasing H₂S content.

   The Strategy-A system determines an environmental severity ranking for pipeline environments in terms of these factors on an interval scale of 1 to 10. A typical dialog box used in Strategy to accept user input is shown in the figure below. The program computes the pH using the acid gas partial pressures and the bicarbonates using relationships given by Crolet. The severity is determined as a function of pH, H₂S and temperature. Effective dehydration, persistence of oil phase and chemical inhibition affect the severity of hydrogen charging and therefore that of HIC and SSC. Dehydration generally works to reduce corrosion and hydrogen charging by minimizing the presence of liquid water in the system. Similarly, both the oil phase and chemical inhibition act as a corrosion resisting barrier between steel and corrosive environment. This step ultimately leads to the determination of an environmental severity factor (ESF).

   
2. Ranking of steels based on material parameters such as composition, micro structure, heat treatment, strength/hardness and material form to determine a material resistance factor (MRF) representing a steel's resistance to HIC, SSC and SOHIC. The Strategy-A system uses a combination of 15 parameters to determine individual material ranks that represent the material's resistance to HIC, SSC and SOHIC. User input to the system can be accessed through the databases (previously stored data) or can be specified interactively through the material dialog box shown below
   

   Heuristic rules gleaned from literature and domain expertise to make judgments about resistance of a specific material to HIC, SOHIC and SSC in a specific environment. Each cracking phenomena has a set of parameters that play a critical role in determining material resistance. For example, for HIC resistance, the following parameters are important:

   • Product form (hot-rolled plate or pipe, forging, casting etc.)
   • Heat treatment (as-rolled, annealed, normalized or quenched tempered)
   • Micro structure (centerline segregation, ferrite/pearlite banding, minor banding, no banding)
   • Plate thickness and yield strength
   • Inclusion morphology
   • De-oxidation practice (Si-Al killed, Si-killed, Ca-treated)
   • Sulfur Very low <0.002%
     Low 0.002 - 0.01%
     Medium 0.011 - 0.02%
     High > 0.02%
   • Phosphorus Low 0.01%
     High >0.1%
   • Carbon Very Low 0.1%
     Low 0.11 - 0.15%
     Medium 0.16 - 0.2%
     High > 0.2%

   Other material compositional factors include Copper, Manganese and Calcium. Copper levels are linked to improved material performance in non-cyanide, medium pH environments. Similarly, high Manganese (> 1.2) can adversely affect the inclusion morphology. Calcium treatment is linked to Ca/S ratio as also the deoxidation practice used in making the steel. Fully killed steels, which can have elongated inclusion clusters can benefit from proper Ca treatment.

   All these compositional elements affect HIC resistance such that increasing levels of these elements typically reduce HIC resistance of both pipeline and plate steels. Parameters affecting SOHIC resistance are,

   • Sulfur content
   • micro structure
   • HAZ hardness
   • operating stress ratio

   Parameters affecting SSC resistance in Strategy are,

   • Weld and base-metal hardness
   • operating stresses
   • minimum operating temperature
3. Normalization of the environmental and material ranks to assess the relative suitability of a given steel to a given environment. In this step, the system compares the extent of corrosive severity of the environment to the resistance of the material from steps 1 and 2. An ESF higher than MRF for any of the three phenomena (HIC, SOHIC or SSC) would indicate a situation where the material will have to be subjected to laboratory evaluation and possibly replaced prior to continued use in the environment.
4. Assessment of reliability of continued use of a specific piece of pipeline equipment and material through determination of a Crack Growth Factor (CGF) on an interval scale of 1-10, determined as a function of the following parameters:
   • Primary and residual stresses
   • Stress relief
   • Toughness
   • flaw size
   • Stress concentration factor
   • Age of equipment
   • Weld joint geometry
   • Shape factor for cracks

The CGF is determined to provide the end-user a vehicle to assess relative crack growth potential between different types of pipeline equipment and to pin-point areas of concern.

The ESF, CGF and the MRF together will allow the user to determine the usability of a material under a given set of conditions. The CGF and the ESF are to be treated as severity factors, meaning lower values for these rankings are preferable. A lower ESF indicates an environment of lower severity. Similarly, a lower CGF indicates a more reliable piece of equipment. The MRF for HIC, SOHIC and SSC determined by Strategy represents material resistance, hence, a larger MRF indicates a less susceptible material. A flow chart depicting the different reasoning modules and their inter-relationships in the Strategy-A system is shown below