| 講演要旨(和文) | | オイルサンドを対象とした岩相モデルを構築した。SAGD法の生産能力は貯留層中の泥岩の影響を受けるとされるが、礫状の泥岩と層状の泥岩では影響の大きさも異なると考えられる。泥岩の形状に応じた岩相の識別が必要である。そこでまず、岩相毎の地震波属性頻度分布から、岩相の識別に有効な地震波属性を選択した。次に、それらの地震波属性を用いてマルチアトリビュート解析を実施し、Probabilistic Neural Networkによる各岩相の出現確率が推定された。最後に、この出現確率を局所的な岩相比率としたSequential Indicator Simulationにより岩相モデルを構築した。岩相モデルは孔隙率および浸透率のモデリングに利用される。岩相識別のために用いた地震波属性の一部につき物理的意味を考察した。こうした貯留層特性と地震波属性の関係の理解が貯留層モデルの信頼性向上のために重要である。 |
|
|
| | 講演要旨(英文) | | A facies modeling was conducted for oil sands where SAGD performance for bitumen production would be affected by mudstone contained in the reservoir. It was essential to distinguish between mud clasts and mud thin layers, which have different impact on production performance. Procedure of the modeling is as follows. Firstly, seismic attributes were investigated for their potential to distinguish the litho-facies by means of histograms of seismic attribute values for each facies. Then, multi-attribute analysis was employed to estimate the facies distribution using the seismic attributes selected in the previous step. Probabilistic Neural Network (PNN) provided probability of each facies at all grid nodes of the model. Finally, Sequential Indicator Simulation was implemented for the estimation of the spacial distribution of each litho-facies, where the PNN probability was used as a local proportion of the facies. The facies model is to be used for the reservoir modeling to estimate the spacial distributions of porosity and permeability. P-wave velocity and instantaneous frequency were discussed for their physical meaning in the reservoir characterization. It should be important for the higher quality of reservoir models to investigate a physical relationship between seismic attributes and reservoir properties. |
|
|
|
| |
|