qfoldit-oilgas

qfoldit-oilgas (corrosion)

Kinetic/empirical model for internal CO2 corrosion of carbon steel oil & gas pipelines, built on the published de Waard-Milliams correlation plus three well-documented secondary effects: protective FeCO3 scale suppression at higher temperature, flow-driven mass-transfer enhancement (with an erosion-corrosion regime at high velocity), and a chloride/ salinity-driven pitting risk indicator.

Read references/model_documentation.md before answering -- it has the full equations, the model's valid range, calibration status, and critical caveats (the base correlation and secondary-effect parameters are generic literature values, not fitted to any specific qFoldIT pipeline/field until real coupon/ER-probe/field data is supplied).

How to handle a request

1. Gather inputs: temperature (C), CO2 partial pressure (bar) -- estimate from total pressure x CO2 mole fraction if not given directly -- flow velocity (m/s), salinity (ppt, default seawater ~35 if offshore/produced-water context implies it), and whether a corrosion inhibitor is in use (and its claimed efficiency, typically 80-95% if properly maintained -- ask rather than assume if not stated).
2. Validate ranges: the de Waard-Milliams correlation's empirical basis is roughly T=5-150C, pCO2=0.1-10 bar. Flag extrapolation outside this rather than presenting it with the same confidence.
3. Run scripts/corrosion_kinetics.py: predict_corrosion_rate(...) for the uniform corrosion rate + pitting risk factor; remaining_life_years(...) if wall thickness is known or asked about.
4. Always separate uniform corrosion rate from pitting risk -- they are different corrosion modes (general metal loss vs. localized perforation risk) and should never be combined into one number or implied to be interchangeable.
5. State calibration status with any numeric forecast: the base correlation is a widely-used published one, but the scale-suppression and flow-enhancement terms are simplified approximations calibrated to generic literature shapes, not to any specific pipeline. If the user has field/coupon corrosion data, use fit_dewaard_offset_from_data to calibrate a site-specific correction factor before forecasting.
6. Remaining-life estimates assume a constant rate -- flag this explicitly; real field conditions (water cut, temperature, inhibitor availability) change over time, so this is a first-order screening number, not a substitute for an inspection program.
7. The flow/mass-transfer factor here is a deliberately simplified proxy, not a reproduction of the full de Waard 1995 Sherwood-type mass-transfer equation (which needs diffusion coefficient, viscosity, and hydraulic diameter -- inputs this skill doesn't collect). It captures the same two qualitative behaviors (mass-transfer-limited at low velocity, erosion-corrosion above an erosional threshold) but should not be presented as numerically equivalent to the full published model. Say so if asked how this compares to NORSOK M-506 or the full de Waard 1995 equation.