Study On Cost Forecasting Methods Of Oilfields Block Based On Costdrivers

With the deepening of the domestic oil field development, many oil blocks come into the mid-late development, the moisture content and drilling difficulty is rising, huge cost investment is needed in order to maintain production levels to compensate for the natural decline, the cost of oil field blocks is rising quickly. Now the main methods of oil block cost prediction is a reference to history and forecasts according to the plan for this year, and it is short of reliability and accuracy. Therefore, looking for suitable cost prediction method which is fit for oil field actual situation and has a higher precision is of great significance to control the cost and improve the economic benefits.Based on the above background, the oil block cost prediction methods are discussed in this paper. First of all, the composition and characteristics of oilfield block cost are analyzed, and on this basis, cost drivers are chosen and cost drivers combined model is established, then cost function is obtained based on cost drivers. Through evaluating the cost prediction method and combining with the characteristics of oilfield block cost prediction, BP neural network prediction method is chosen from many methods. Finally, taking oilfield blocks lift system cost prediction as an example, and the prediction results are compared with the regression forecast method, exponential smoothing method, moving average method, and it is concluded that the BP neural network prediction is beneficial to improve the prediction accuracy of oil block cost. In the specific application, it also should strengthen the basic work of cost drivers data, perfect the information management system, and further improve the level of oilfield block cost prediction.