The corrosion rate cannot be specified for generic cases, because the error/deviation can be significant. Regional studies over a period of time have generated certain data used for estimation of possible corrosion rate range, without guaranteeing those figures for corrosion rate predictions of a structure, pipe of plant equipment. You might have noticed that the environmental conditions tend to be variable and tend to defy any strict prediction. You could talk to relevant people in your area (a refinery, a gas distribution centre or the like) and ask what conservative figures they might use for corrosion rate predictions. Use your experience to predict what are your chances of avoiding holydays in the pipe coating, accidental break down of that coating applied on the pipe, likely to generate accelerated localised corrosion. Yes, this is practical theory. To answer you specific question, in accordance with the environmental conditions you have assumed (desert, dry, etc..);- you don't need any coating, because the carbon steel to corrode need moisture/water, oxygen and temperature. Since you have removed the moisture, there is no corrosion. Easy done, see? Maybe not, perhaps is something wrong with the assumption...
cheers,
gr2vessels