It all depends on the composition of the fluids in the pipe ... there are a variety of models out there for trying to predict corrosion rates but I'm not sure how accurate they are.

You mention you have minimum thickness readings - is this new pipe and you have baseline data that you took or is this old pipe and you have taken readings of current condition?

Generally, you have to have at least two sets of readings taken at the same monitoring point in order to calculate the actual real corrosion rate at those locations. It is very inaccurate to use a single set of readings since you don't actually know what the starting thickness of the pipe was.

Anyway, I think it would be helpful to post some information on the fluid composition to help folks try and provide you some guidance.

the simplest way is to divide the thickness loss by the time of operation, but as rneill said the more data you have the more accurate will be the determination of the corrosion rate. for example if you have just one reading after then years of operation you'll have a mean corrosion rate over a 10 ys period but is possibile that the corrosion has taken place only in the last years and in this case the corrosion rate will be much higher..
there are several standard like the API 581 for risk based inspection that gives you a methodology to assess the likelihood of failure and corrosion rate based on NDT and literature data:

 http://www.techstreet.com/cgi-bin/detail?doc_no=API%7CRP_581&product_id=1579565

or the API 579

http://catalog.asme.org/Codes/PrintBook/FFS1_2007_FitnessforService.cfm

if you need corrosion data :

http://www.corrosionist.com/isocorrosion_curves.htm

hope this help

S

Corrosion & Rust Prevention Control
 

www.hse.gov.uk/research/rrpdf/rr016.pdf

Steve Jones
Materials & Corrosion Engineer
http://www.oilandgaspeople.com/cv/11499664
http://www.linkedin.com/pub/8/83b/b04