Corrosion Prediction
by: Gordon A. Aaker, P.E.
The following SAMPLE CALCULATION may be
used as a guideline to predict the corrosion
rate from the effects of CO2
1. Design Conditions:
Operating Condition = 245 psi (16.33 Bars),
Mol % CO2 = 4.0%,
Partial Pressure CO2 = 9.8 psi (.6533 Bars),
Temperature = 120 degrees F (48 degrees C).
2. Corrosion Effects
The corrosion of carbon steel in a wet CO2 environment can
lead to iron carbonate as a reaction product. Although recent work
suggests that an iron carbide matrix is first produced on the surface of
a corroding steel surface, a carbonate scale, which will protect the
underlying metal, can often be formed. The formation and protectiveness
of such a scale depends on a number of factors such as the solubility of
iron carbonate (which will vary with pH and the presence of other salts),
the rate of reaction of the underlying steel and the surface quality
(roughness/cleanliness). In addition, the scale persistency will be
sensitive to turbulence and may grow to such a thickness that it cracks
and spalls. Current practice is to not rely on such carbonate scale to
give continuous protection over the full length of a hydrocarbon flow
line or export line. Particularly in regions of high flow or at weld
seams, scale breakdown can lead to a rapid rates of localized corrosion
(Mesa Attack). Although a carbonate scale may protect the majority of a
line, the corrosion rate used for design calculations must be based on
the de Waard and Milliams equation. However, at temperatures of 60°C
or more the scale is considered stable and protective enough to allow
modification of the de Waard an Milliams equation by a multiplying
factor,
where V is the adjusted corrosion rate, F(s) is the factor for scaling
and v(dWM) is the corrosion rate calculated from the original de Waard
and Milliams equation.
Log F(s) = (2500/(T+273)) - 7.5
Log F(s) = (2500/(48°C+273)) - 7.5
Log F(s) = (2500/321) - 7.5
Log F(s) = 7.788 - 7.50 Log
F(s) = 0.2882
F(s) = 1.941
Since the is temperature is less than 60° C use 1.0.
If the correction for high temperature scaling is made then no allowance
should be made for pH correction. The pH correction factor is for bare
metal surface and it is currently unclear what effect pH will have on the
corrosion rate of a surface that is protected by a high temperature
scale. The original de Waard and Milliams work was carried out at low
partial pressures of CO2 (< 1 bar). The fugacity of a gas
species is effectively the activity of a gas species and for ideal gases
or gases at low pressures this is equal to the partial pressure. At
higher pressures, because of non-ideality, the fugacity or activity of
CO2 is less than its partial pressure and this should be used
in the calculation of corrosion rates. It is possible to calculate the
CO2 fugacity (f CO2) and use it in the de Waard and
Milliams equation corrected for partial pressure:
where f CO2 is the CO2 fugacity, (a) is the
fugacity coefficient and ppCO2 is the partial pressure of
CO2
For operating pressure of 245 psi
ppCO2 = Pressure( % CO2)
ppCO2 = 245 psi (4.0%)(.01)
ppCO2 = 9.8 psi or 0.6533 Bars
Log(a) = P total (0.0031 - 1.4/(T + 273))
Log(a) = (245/15)Bars(0.0031 - 1.4/(68°C +273))
Log(a) = (16.33Bars)(0.0031 - 1.4/321)
Log(a) = (16.33Bars)(0.0031 - .0044)
Log(a) = (16.33Bars)(-0.0013)
Log(a) = -0.0212
(a) =.9523
then, the calculated corrected partial pressure is:
f CO2 = (a)(ppCO2)
f CO2 = (0.9523)(0.6533 Bars)
f CO2 = 0.6221
the corrosion rate v(dWM) is given by:
Log(v) = 5.8 - 1710/(t + 273) + 0.67 Log(f CO2)
Log(v) = 5.8 - 1710/(48 + 273) + 0.67Log(0.6221)
Log(v) = 5.8 - 1710/(321) + 0.67(-.2061)
Log(v) = 5.8 - 5.32 + - 0.1381
Log(v) = 0.3419
(v) = 2.19 mm and the adjusted corrosion rate V is:
V = F(s) v(dWM)
V = (1.0) ( 2.19 mm)
V = 2.19 mm per year.
The use of inhibitors can be 90% effective, therefore, with inhibitors,
V= 0.219 mm per year. For 20 Years:
V = 0.219 mm(20 yrs)
V = 4.3 mm = 0.17244 inches
The above prediction is recognized as being very conservative and
usually is modified
V = 0.177 inches x 0.60 = 0.120 inches
Required minimum corrosion allowance for 20 years is 0.102 inches
3. Velocity Effects
It is known that in mixed phase conditions without the presence of
solids that a slug flow condition readily damages any scale formed. As a
result, where scale damage has occurred carbon steel corrosion rates
would be at least equal to those predicted if no iron carbonate scale is
formed. When scale is damaged pitting corrosion will take place. For
systems, which are primarily liquid phase, and without solids,
superficial flow velocities exceeding approximately 2.7 meters per second
cause scale damage and corrosion similar to the slug flow condition. If
solids are present, any protective scale is readily removed and corrosion
rates will be at least as great as predicted if no iron carbonate scale
is formed.