热采井用固井水泥石养护方法及力学性能研究

如何模拟稠油热采井井下环境,对评价稠油井用固井水泥石能否满足热采工况的需求具有重要意义。然而,以往用马弗炉超高温干燥条件养护水泥石的方法,与稠油热采井下的超高温水蒸气环境、水泥石受套管和地层约束等实际情况不符。为此,提出了一套可模拟稠油热采井下超高温水蒸气环境的固井水泥石养护装置及方法,并对比研究了超高温干燥与水蒸气条件下,试样尺寸、加热速率对固井水泥石抗压强度和完整性的影响。研究结果表明,小尺寸水泥石试样、低升温速率和水蒸气环境有利于保证在加热过程中试样受热更均匀,从而有利于缓解因受热不均导致开裂的现象,进而维持水泥石更高的抗压强度及完整性。因此,通过该研究结果,建议室内实验模拟方法应充分考虑实验条件对模拟结果的影响;同时,基于此研究结果,建议在实际生产过程中,可适当优化注蒸汽的工艺过程及参数,以降低注蒸汽过程对固井水泥石的加热速率、减少对水泥石的不利影响。

Nuclear magnetic resonance T_2 spectrum:multifractal characteristics and pore structure evaluation

Pore structure characteristics are important to oil and gas exploration in complex low-permeability reservoirs. Using multifractal theory and nuclear magnetic resonance （NMR）, we studied the pore structure of low-permeability sandstone rocks from the 4th Member （Es4） of the Shahejie Formation in the south slope of the Dongying Sag. We used the existing pore structure data from petrophysics, core slices, and mercury injection tests to classify the pore structure into three categories and five subcategories. Then, the T2 spectra of samples with different pore structures were interpolated, and the one- and three-dimensional fractal dimensions and the multifractal spectrum were obtained. Parameters a （intensity of singularity） andf（a） （density of distribution） were extracted from the multifractal spectra. The differences in the three fractal dimensions suggest that the pore structure types correlate with a andf（a）. The results calculated based on the multifractal spectrum is consistent with that of the core slices and mercury injection. Finally, the proposed method was applied to an actual logging profile to evaluate the pore structure of low-permeability sandstone reservoirs.

Marchenko imaging based on self-adaptive traveltime updating

Marchenko imaging obtains the subsurface reflectors using one-way Green’s functions,which are retrieved by solving the Marchenko equation.This method generates an image that is free of spurious artifacts due to internal multiples.The Marchenko imaging method is a target-oriented technique;thus,it can image a user specified area.In the traditional Marchenko method,an accurate velocity model is critical for estimating direct waves from imaging points to the surface.An error in the velocity model results in the inaccurate estimation of direct waves.In turn,this leads to errors in computation of one-way Green’s functions,which then affects the final Marchenko images.To solve this problem,in this paper,we propose a self-adaptive traveltime updating technique based on the principle of equal traveltime to improve the Marchenko imaging method.The proposed method calculates the time shift of direct waves caused by the error in the velocity model,and corrects the wrong direct wave according to the time shift and reconstructs the correct Green’s functions.The proposed method improves the results of imaging using an inaccurate velocity model.By comparing the results from traditional Marchenko and the new method using synthetic data experiments,we demonstrated that the adaptive traveltime updating Marchenko imaging method could restore the image of geological structures to their true positions.

Seismic multiple attenuation based on improved U-Net

Eff ective attenuation of seismic multiples is a crucial step in the seismic data processing workfl ow.Despite the existence of various methods for multiple attenuation,challenges persist,such as incomplete attenuation and high computational requirements,particularly in complex geological conditions.Conventional multiple attenuation methods rely on prior geological information and involve extensive computations.Using deep neural networks for multiple attenuation can effectively reduce manual labor costs while improving the efficiency of multiple suppression.This study proposes an improved U-net-based method for multiple attenuation.The conventional U-net serves as the primary network,incorporating an attentional local contrast module to effectively process detailed information in seismic data.Emphasis is placed on distinguishing between seismic multiples and primaries.The improved network is trained using seismic data containing both multiples and primaries as input and seismic data containing only primaries as output.The eff ectiveness and stability of the proposed method in multiple attenuation are validated using two horizontal layered velocity models and the Sigsbee2B velocity model.Transfer learning is employed to endow the trained model with the capability to suppress multiples across seismic exploration areas,eff ectively improving multiple attenuation effi ciency.