Reservoir Prediction Technology And Application Using Kriging Method

In order to resolve the problem that the low precision in reservoir prediction that because of the low resolution the noise and the reservoir anisotropy in Erduos daniudi Gas Field, kriging method is used to predict reservoir in this area. In this area, factorial kriging is used to delete the noise in the attribute maps and Bayesian kriging is used to predict the reservoir and make model of the reservoir. At the same time, The advantages and disadvantages of the traditional ways for oil prediction are discussed again in this article. The effect of this method has been proved by analyzing the theory and testing data in this area.There are three parts of this article:(1)The analysis of reservoir prediction in Daniudi. First the article analyze characterize of the Daniudi Gas Field and then the paper makes out the process of applying factorial kriging and bayesian kriging method. Last, it makes a conclusion about the current reservoir prediction methods and shows some advices after discussing the advantages and disadvantages of the traditional methods ;(2) Factorial kriging is used to delete the noise in the attribute maps. Factorial Kriging is a geostatistical approach to isolating and extracting features related to the modeled variogram. In this case, the footprint in the attribute map can be isolated and removed very simply; (3) Bayesian kriging is used to predict the reservoir and make model of the reservoir. The traditional kriging-Ordinary kriging only use the well data to make prediction, and the seismic data can't be used effectively. Obviously, the reservoir won't be predicted subtly by this way. Bayesian Kriging can combine the well data with seismic data to predict the reservoir. This method can resolve the problem that how can combine the well data with seismic data to predict reservoir. And these methods have good effect on reservoir prediction and geostatistical modeling and inversion.The paper makes out the process of applying factorial kriging and bayesian kriging method. There are good effects on filtering the noise of the attribute map. Reservior and parameter's distribution maps are achieved. And the modeling and inversion of the reservoir using well data and seismic data also have a good effect. Finally, it makes a conclusion about the two kriging methods and shows some advices. This technology both has a profound theory value and gets a very important practical value. There must be a well outlook on this method in the future.