If it is vapor phase, the velocity may not be a big issue up to 250 fps when there are no solids present. In the multi or liquid phase, the most variable parameter in corrosion-erosion with velocity/wall shear stress will be oxygen contents. However when O2 is below 2-3 ppb, normally the corrosion rate is not changed.
Most simple calculation models of CO2 corrosion do not have the input data with velocity. So my recommendation is to apply only the erosion velocity and the recommended velocity per pipe size and phase in the process engineering guidance.

If you are interesting in the detail technology for the erosion (not corrosion-erosion mode) by velocity and particles, please find the following models.

- AEA Technology’s Sandman (Sman 39)
- University of Tulsa’s (E/CRC) Sand Production Pipe Saver (SPPS)
- DNV’s RP-O501
- Salama’s OTC 8898