Corrosion Rate Assessement

[ChihebG]

Experts,

I have calculated the corrosion rate of CO2 corrosion of free water contaminated with CO2 (piping) (CO%=0.35%; T 38 °C; P=29 barg) with 3 corrosion prediction models:
DeWaard 95
NORSOK M-506
Cassandra model(BP)

I have found:
DeWaard 95 :0.06 mm/yr
NORSOK M-506 :0.1 mm/yr
Cassandra model(BP):0.56 mm/yr

What is the corrosion rate that I have to choose to define the appropriate strategy of corrosion control knowing the NORSOK model seems to be very conservative ?

 

[SJones]

If you are part of the operator's organisation, you will research the theory and development of each of the models and arrive at a decision regarding the one model that best fits the design configuration and the end user's beliefs about CO2 corrosion. You will then stick with that model.

If you are part of the engineering contractor's organisation, you will do much as you have done here and ask the end user client to direct you. You will then stick with that model.

You can, of course, go to another free model: FREECORP from Ohio University, and see what that comes up with, but that will give you a fourth value over which to ponder.

You can research NORSOK M-506 using NACE Corrosion 2005, Paper 05551, and Cassandra using Paper 05552. If you research the de Waard - Milliams model development, you will note that it was modified further around 2002. It would also be interesting to note that, further to the BP statement regarding no further development of Cassandra in Paper 05552, they completely ditched it in 2010 and replaced it with a proprietary 'bells & whistles' version of FREECORP bearing little resemblance.

My only advice to people who attempt to use model predictions to two or more decimal places in attempting to justify carbon steel is: how lucky do you feel? Meaning, if you need to use inhibitor how reliable is the inhibition activity as a whole; if you think that inspection will keep an eye on things, what coverage, how frequently, and how much will it cost?

As the IFE Guideline for CO2 Corrosion Prediction that I gave a link for in your other thread states:

"The issue is how well the design corrosion rate predicts the corrosion rate eventually measured in the field."

So, how lucky do you feel with regards to any of those three models coming anywhere near what gets observed when a carbon steel system actually goes into operation?

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/pub/8/83b/b04

All answers are personal opinions only and are in no way connected with any employer.

 

[ChihebG]

Thank you Steve!