The prediction models of CO     2     corrosion rate are summarized, and the corrosion mechanism, corrosion morphology,      and influence factors of each model are compared and analyzed in combination with CO     2     corrosion experiments under actual      production conditions in the Bohai oil field. There are three types of CO     2     corrosion prediction models around the world. They are      the empirical prediction model, semi-empirical prediction model and the mechanism prediction model. The empirical model is      built considering less corrosion mechanism but more experimental data and field data. And the Norsok M-506 model is one of the      most famous empirical models. The semi-empirical model is established by fitting experimental data and field data based on the      dynamics process and the medium transmission process. The DWM, DW91, DW93, and DW95 are well-known semi-empirical      prediction models for calculating the CO     2     corrosion rate. The mechanism model uses dynamic formula to calculate the CO     2      corrosion rate, which has a clear physical meaning and is less integrated with experimental data and field data. The Nesic model      is most classic mechanism model to predict the CO     2     corrosion rate. Based on a semi-empirical and semi-mechanical model, the      CO     2     corrosion rate prediction model is built in this paper fully considering the effects of temperature, CO     2     partial pressure, flow      rate, CO     2     corrosion product film, and the CO     2     corrosion rate correction factor. Besides, a CO     2     corrosion experiment under actual      production conditions in Bohai oilfield is conducted in order to let the CO     2     corrosion rate prediction model more suitable for the      Bohai oilfield. The comparative study of the model established in this paper with DWM, ECE, and BP semi-empirical models re     vealed that this paper’s developed model has the smallest average relative error and is similar to the actual rate of CO     2     corrosion      in the Bohai oilfield, confirming that the CO     2     corrosion rate prediction model built in this paper has great prediction effects and      is more accuracy. Both the results of the corrosion experiment and model predictions in this paper have revealed that N80 pipe      has the worst resistance to CO     2     corrosion and the 13Cr pipe has the best resistance to CO     2     corrosion. The ranking of four kinds of      pipes for CO     2     corrosion resistance is N80 < 1Cr < 3Cr < 13Cr. The wellbore corrosion rate distribution forecast contrasts the CO     2      corrosion rate of various pipe materials, which provides a theoretical framework for the optimization of oil production and casing    

pipes in the Bohai oilfield.