Study Of Prediction For Effect Of Major Technologies And Its Planning Methods At Sandstone Reservoirs

After high water cut period, potential wells that adopt augmented injection and increase production measures become fewer, and economic benefit declines year by year, oil field confronts many problems such as high water cut of main layers, production rate decreases progressively, remaining oil becomes little, exploitation changes into difficult and so on. It's necessary for fields to predict the effects of main measures to evaluate their potentials and optimize measures'structure, criteria can be provided for measures'design and theory can be given for reasonable measures'scheme, so the effects of measures can be developed.The research raises a method that correcting geologic parameters with performance data under two situations: layering test data known and layering test data unknown. Measure layer can be incorporated by layers implemented measures with their little layers geology parameters and production parameters weighting, and other layers incorporate to non-measure layer, so the actual geology model can be changed into ideal geology model containing measure layer and non-measure layer.On the basis of the ideal geology model, effects of five kinds of main measures which are oil wells fracturing, water shutoff, water wells fracturing, water wells acidifying and water wells'adjusting sections should be predicted. The analysis model of single well measure effect predicting that can meet the need of petroleum engineering planning is built by means of filtrational resistance. On the basis of data of wells which have been taken measures, factors that influence the effects of measure taken on the wells of no.1 and no.2 reservoirs are analyzed, samples bank is built. And models for predicting the measures'effects are built by means of multiple element linear regression and artificial neural network. With the models effects of measures can be evaluated under different technology and reservoir types conditions, and then the statistics relations can be used to predict measures'effects.With the predicting results gained form the two methods depicted above, measure work load is optimized and candidate wells are selected, so the whole economy beneficial can be raised utmost. Two kinds of models of measure planning are built under cost controlling and development index controlling preconditions, in this way, solving the measure problems is changed into solving the whole optimization with limited conditions which can be solved by way of genetic algorithm.On the basis of research above, software that can be used to calculate measure planning of reservoir engineering is compiled. It can realize the integration of parameters preprocessing, effects predicting and measure planning , effects predicting dynamically and currently is also achieved.