Search

Defining the boundaries, thicknesses and sedimentary facies of fluvial reservoirs (sand bodies) is critical for predicting hydrocarbon volumes, designing schemes for petroleum exploration and development and improving oil recovery. Most reservoirs contain thick and thin sand bodies at the same intervals, while the amplitude values of seismic data usually highlight sand bodies near the 1/4 wavelength for the tuning phenomena. Hence, the application of spectral decomposition to seismic attributes and the combination of multiple frequency-decomposed (spectral-decomposed) seismic attributes have gained increasing attention for the readjustment of tuning thickness to predict sand bodies of various thicknesses. However, the popular method of red-green-blue blending is a simple linear combination of three frequency-decomposed seismic attributes that qualitatively analyzes the sand thickness without well-log interpretation. This research proposes machine learning fusion as a new nonlinear method for fusing high-, middle-, and low-frequency seismic attributes. This method uses machine learning to link well-log interpretation and multiple-frequency seismic attributes for the quantitative prediction of sand thickness, which is important for development work in a mature field. Test results of the conceptual model and the real case indicate that the predicted sand thickness after fusing multiple frequency-decomposed seismic attributes is approximately in line with the actual thickness (correlations between 80 and 90%). Combined with the coherence attribute and the red-green-blue blending results, the distributions and histories of sedimentary facies are analyzed based on the predicted sand thickness and well data. The results suggest that the proposed method can effectively readjust the tuning thickness and improve the resolution of seismic interpretation. This method is a potentially effective technique to characterize the sand thickness and sedimentary facies in other fields using a similar geological setting and dataset.

The reservoir architecture of a meandering river deposition can indicate the heterogeneity of the subsurface reservoirs and deposition and evolution of the paleo-river. Although outcrop study has been useful in characterizing reservoir architecture, the study of the subsurface reservoir is still a challenge due to its complicated reservoir geometry and forming process. In this work, a mature oil field in Bohai Bay Basin, eastern China was investigated to understand the reservoir architecture and evolution of a Neogene meandering belt using drilling cores, well logging, 3D seismic data, and oil production data. In the study area, five types of litho-facies were observed indicating the specific hydrodynamic conditions. The results obtained two types of point bar combination models focused on (1) the point bars migration in the same direction at equal elevation and (2) erosion and superimposition of the multi-phase point bars. Abandoned channels in study area are classified into two types: Type Ⅰ is the “Early and long-term abandoned channel” and Type Ⅱ is the “Late and short-term abandoned channel”. Well log facies of Type Ⅰ presents a low sandstone/mudstone ratio while that of Type Ⅱ has a high sandstone/mudstone ratio. Both models exhibited the weakening of the seismic reflection amplitude and changes in the waveform. Based on the well log curves, stratal slicing, and seismic facies analysis, single point bars are interpreted more accurately. The meandering river in the study area, which was previously interpreted as a large single point bar, was comprehended as a meandering belt formed by a multi-phase point bar complex with three single point bars in the meander. Erosion and overlapping can be observed between the first and second phases of the point bars. In modern river deposition, similar meander reservoir architecture and its evolution can be observed. Additionally, reservoir engineering analysis showed that there are flow barriers between point bars of different phases. The reservoir architecture of a meandering belt was characterized and the depositional evolution of the paleo-river was reconstructed. The methods proposed is significant for qualitative interpretations of reservoirs such as morphology channel recognition, reservoir connectivity analysis, and lithology and interlayer prediction.

This paper studied the impact of the photoelectric effect of X-ray density logging in order to improve the accuracy of X-ray measurement density. Firstly, the laws governing the changes of the Compton scattering and photoelectric effect of X-rays in the formation were analyzed based on the principle of X-ray attenuation in the formation and using the XCOM program. Secondly, the impact of photoelectric effect on X-ray density logging was studied on the basis thereof. Thirdly, a method for correcting photoelectric effect by calculating the formation bulk density using the counts in both lithology window and density window was proposed. Finally, the proposed method was verified through MCNP simulation. The results indicate that photoelectric effect was an important factor affecting the accuracy of X-ray density measurement, and the lower the X-ray energy was, the greater the impact of photoelectric effect on X-ray density measurement would be and the lower the accuracy of density measurement would be. As the γ-rays released from a Cs-137 source were at a high energy level, the impact of photoelectric effect was basically negligible when the formation bulk density was calculated using the counts in density window; on the contrary, the X-rays released from the X-ray source was at a low energy level, the impact of photoelectric effect could not be neglected when the formation bulk density was measured using X-ray. The method for calculating the formation bulk density proposed in this paper can effectively eliminate the impact of photoelectric effect and reach up to a measurement accuracy level of about 0.01g/cm3, providing a theoretical foundation for accurate density measurement during X-ray density logging.

Effective identification of lithology using well logs is one of the most important steps for reservoir characterization. A lot of methods have been developed to identify lithology automatically by analyzing the value or patterns of well logs. However, the response characteristics of log curves for some different lithologies are similar and indistinguishable. As a result, we may obtain results which do not follow the general geologic rules if we only use values or patterns of well logs as the criteria to identify lithology. In this paper, we propose an alternative automatic method to identify lithology by integrating both well logs and sedimentary patterns. First, we obtain the probability of lithology by applying the artificial neural networks (ANN) to the well logs. Then, we obtain the vertical combination patterns of different lithologies by applying a probabilistic statistical method to the cores. Finally, we produce an integrated lithology probability by combining the lithology probabilities calculated using ANN and lithology probabilities computed using sedimentary pattern. We validated our method by applying it to a real case study from China. The results indicate that the accuracy of lithology identification using the proposed method is much higher (91%) than that using neural networks method (83%).