Font Size: a A A

Study On Forecasting Method Of Oil - Gas Two - Phase Capacity Based On The Combination Of Surveying Well Information

Posted on:2016-12-30Degree:MasterType:Thesis
Country:ChinaCandidate:L MaFull Text:PDF
GTID:2270330470952825Subject:Earth Exploration and Information Technology
Abstract/Summary:PDF Full Text Request
Some factors resulted in incorrect prediction of oil and gas production in Yingdong oil-gas field in Qaidam Basin, such as the long oil-bearing well section, the thickness of a single layer, a large number of layers and faults, the complex relation of oil and gas,which adds a lot of difficult to the work of production. For that situation, the methods of gas-oil division and ratio calculation have been researched in this paper, as well as the application of two-phase productivity prediction of oil and gas. And it will make up the blank research of two-phase productivity prediction, and will finally optimize the production horizon, boost productivity construction. In the methods of gas-oil division and ratio calculation research, I developed the spider diagram to distinguish oil and gas qualitatively, and then found the positive correlation of gas-oil ratio (GOR) and the ratio of different component area in spider diagram, and thus can be fitted into a quantitative calculation model. In the research of productivity prediction, I improved Vogel inflow-performance-relationship (IPR) by setting a negative increment hands the oil permeability reduce due to free gas, which greatly reduces the fitting error and gives it a wider applicability, using in the well with both oil and free gas, and it has been used in this paper for the conversion of open-flow capacity and real production. Next, the BP neural network is used to predicting open-flow capacity. Before that, the input layer and output layer of BP neural network have been determined through the analysis of productivity effects. And then the production and be figured out associated with GOR. By integrated all the steps, a complete set of production prediction application can be developed., which have been used in8well’s production prediction in this paper and the production relative error is less than10%, as well as the GOR relative error is within20%, and this is enough to achieve the basic requirement of judging whether there is a yield and gas or oil. So the set of methods are proved to be feasible.
Keywords/Search Tags:Qaidam Basin, productivity prediction, inflow-performance-relationship, gas-oil-ratio, BP neural network
PDF Full Text Request
Related items