Two phase flow applications in petroleum industry are so widespread. It is a fact that UBD precise bottomhole pressure maintenance ascertains UBD success. UBD hydraulics design, especially for inclined trajectories, is a real challenge. This is greatly dependent on the pressure drop in the annulus. Two phase flow through annulus is an ambiguous area of study to evaluate the bottomhole pressure. Two phase flow correlations on which most of UBD simulators based on over predict and also make extrapolation risky. Although Mechanistic approaches increase the frequency for designing two phase flow systems in pipes, modeling them through annulus by using the hydraulic diameter concept is not so successful. For this reason, their corresponding errors are not small. Therefore, in this paper, Artificial Neural Network is made use of to evaluate BHP in the inclined annulus using two major Iranian Oil Fields. To compare BHP found by neural network, Naseri et al mechanistic model which is a popular mechanistic model for these two fields is applied. ANN shows to perform much better than Naseri et al mechanistic model. The results show that neural network can estimate bottomhole pressure with an error of less than 20%. This proves that in case of existence of measured BHP while under balanced drilling, it is worth to use ANN to simulate BHP rather than mechanistic modeling or correlations. ANN is highly shown to be useful for solving the non-straightforward problem of two phase flow in annulus. Few jobs have been done to prove the superiority of ANN to mechanistic modeling and correlations in terms of pressure prediction especially in under balanced drilling.

No abstract

No abstract

No abstract

No abstract