J. Basic. Appl. Sci. Res., 2(6)5570-5575, 2012 © 2012, TextRoad Publication ISSN 2090-4304 Journal of Basic and Applied Scientific Research www.textroad.com *Corresponding Author: Payam Hooshmand, Department of energy engineering , Science and Research branch, Islamic Azad university, Tehran, Iran. Stuck Drill Pipe Prediction with Networks Neural in Maroon Field Mostafa Moradi Nezhad 1 , Siavash Ashoori 2 , Payam Hooshmand 3* , Mojtaba Mirzaee 4 1 Iran National Drilling Company 2 Petrolume University of Technology , Ahvaz , Iran 3, 4 Department of Energy and Environment, Science and Research Branch, Islamic Azad University, Tehran, Iran ABSTRACT Drilling practice abounds with a number of problems and an efficient way to deal with problems in this area is a key to continuity of the process. Through careful consideration of some important points and in time action during such practices it is practical to avoid some of these problems or at least lessen them to a desirable point. One of a kind of this problem in drilling practice is pipe sticking in the well. A number of factors are associated with this phenomenon. Today experts and engineers rely mostly on utilizing the old methods which is experimental way to estimate the condition on which such sticking occur. In recent years, Artificial Neural Networks has gain a lot of recognition and its highly functional merits prompt us to use this method in order to estimate pipe sticking in drilling industry. In this research feed forward neural networks and back propagation network training was employed to predict sticking related to pressure difference, well narrowness (Mobile and Chili circulation), weak hydraulic of drilling mud, weak geology of drilling mud, non regulated drilling line , with geological effects were estimated during drilling. The obtained results suggest that neural networks have the potential to predict efficiently drill string sticking in different cases and they proved effective enough to deal with problems in this area. KEY WORDS : drilling - Neural networks - drill string sticking. INTRODUCTION Stuck pipe can be defined in a concise statement as “inside-well force which obstructs the outlet of tubes from well”. In oil and gas drilling, drill string is always accompanied with drilling. As soon as sticking occur many efforts initiates to take them out. Engineers utilize a wide variety of known methods including increased upward strength, increased upward weight and etc. so that they are able release drill string which is stocked inside well yet these methods proved badly chosen and costly, however in most cases they lead to release of drilling line. Drill string sticking turn out to be more and more consequential when due to any possible reason it becomes difficult to take out drill string of the well. In this case the only alternative will be cutting of drilling line inside the well and in the worst case the blockage of the stock and drilling line can occur in this case drilling a secondary rout and changing of drilling plan and extra expenditure happen to be inevitable.[1] This phenomenon is more important in marine practices where pipe sticking will increase about 30% of the whole developmental process of such wells which is highly pricey.[2] Therefore, any attempt to lessen drilling line pipe sticking considered essential in reducing the cost of drilling. Research has conducted on pipe sticking since 1950. In 1985, Kingsborough and Hemp King [3] handled their task through ecstatic analysis of pipe sticking based of drilling parameters. Their work focused on comparing and contrasting stocked and non stocked wells and parameters of both examples examined and subsequent drilling plans carried out through careful consideration of non stocked wells. Kingsborough and Hemp King examined 221 parameters in 131 stocked wells in Mexican well and predicted possibility of sticking in adjacent wells. In 1994, Biegler and Kuhn [4] provided a data collection for 22 drilling parameters in 73 non stocked wells in Mexican gulf and did the same task for 54 stocked well and scrutinized the problem. They are considered as pioneers for primary analysis in this area and conducted their task through integration of drilling variables. They not only predicted the possibility of stuck but also were able to determine the mechanism by which it occurs. In 1994, Howard and Glover [5] improved the existing predictability models in pipe sticking. They carried out their task in Mexican gulf through careful examination of 100 wells. The models they were applying prevented pipe sticking or they could help as a means to release. An artificial neural network has recently been utilizing to predict sticking which is based on pressure difference in Mexican gulf. [6] Drill string sticking is commonly divided into two general categories. First is mechanical sticking and next is pressure difference sticking. During stick resulting from pressure difference tube upward and downward circulation is 5570