Study On The Exploration Planning Of Xinjiang Oil Field Based On Reserve Forecasting

Reserve is the basic resource for oil companies. In all levels of oil reserve, for oilcompanies, proved reserves is the most practical significance. Exploration is the only way toget proved reserves and it is an activity with high investment and great risk. So, it is veryimportant for oil companies to arrange exploration activities reasonable.In recent years, Xinjiang Oilfield has a rapid development. But, in exploration area, thecompany has many problems, such as the low proven degree and the shortage of reserveresources, which limit the development of the company. Based on the actual explorationsituation of Xinjiang Oilfield, this paper analyses the related issues in exploration area and thesource of proved reserves at first. Then, using time series theory, the relationship betweenproved reserves and the three exploration methods used in Xinjiang Oilfield is analyzed, andthe VAR model is built which includes2D seismic,3D seismic, drilling footage and provedreserves. Using difference decomposition, the contribution of the three exploration methods toproved reserves is estimated, and finally the proved reserves forecast model is built. Based onthe actual significance of the sunk costs, exploration cost is regarded as Markov chain. Thestate space and the steady state probability are obtained. Then the value and probability ofexploration costs in different stage are achieved. Based on the actual constraints and theStrategic objectives of future exploration in Xinjiang Oilfield, the dynamic ProgrammingModel about proved reserves and exploration costs is built. Using the model, the strategy setbased on different targets is obtained. Then the exploration work, exploration costs andproved reserves using different exploration method from2013to2015is achieved. The statetransition of Markov chain is used to get state transition matrix in the different costs state. TheMaximum net cash flow is regard as target and linear programming is introduced. Finally, theoptimal strategy which can get the maximum return in different costs state is obtained.