Formation Mechanism And Distribution Of High Quality Conglomerate Reservoirs In The Baikouquan Formation,Junggar Basin

Fan delta reservoir in the Baikouquan Formation,Mahu Sag,Junggar Basin is abundant in hydrocarbon resources.The formation mechanism and distribution of excellent conglomerate reservoirs are the bottlenecks restricting the exploration and development of Triassic conglomerate reservoirs in the Mahu Sag.Due to the influence of tectonism,sedimentation and diagenesis,the conglomerate reservoirs are superposed in multiple stages,and different lithofacies have underwent different diagenesis.It is vital to clarify the variation in reservoir quality and determine the favorable interval thus enhancing oil recovery.Taking fan delta reservoir in the Baikouquan Formation in Ma131 well block of Mabei Oilfield,Mahu Sag as an example,this paper integrates seismic,wireline logs,core analysis and laboratory data,as well as machine learning methods,to elucidate the distribution and controlling factors of channels and fan-shaped lobes in the fan delta reservoir,reveal the control effect of tectonism,sedimentation and diagenesis on conglomerate reservoir quality,and predict the favorable area.It lays the foundation for later well location deployment and development plan formulation.Through this research,the following conclusions can be drawn:The fifth-order reservoir architecture of underwater distributary channels and lobes in lacustrine fan delta reservoirs and their controlling factors are determined.Based on the genesis of conglomerate body and their lateral contact relationship,nine forward conceptual models are designed to provide a prior understanding for the lateral boundary recognition of fan delta reservoir architecture.The results showed that the waveform response characteristics of 45 Hz and 60 Hz frequency division seismic profiles can be used to identify lateral contact relationship between wells.A total of 4 channel belts were identified in the northern part of the study area.3 lobe belts were identified in the lower submember of member 2,and 3 channel belts were identified in the upper submember of member 2 in the eastern part of the study area.The thicknesses of the channel belt in the study area are 2.8 m～12.6 m,with an average value of 6.3 m.The widths of the channel belts vary between 900 m and 3050 m,with an average value of 1884 m,and are positively correlated with the thicknesses.The thicknesses of the lobe belts are 4.5 m～13.8 m and the long axes are 3.5 km～7.3 km.The macroscopic distribution of reservoir architecture is controlled by palaeogeomorphology,and the accommodation space and sediment supply also play an important role in the distribution of reservoir architecture.The effect of tectonism,sedimentation and diagenesis on reservoir quality of conglomerate reservoir is revealed.The contemporaneous reverse fault controls the distribution boundary of fan delta sedimentary facies belts and differential dissolution.The sedimentary process has a remarkable impact on the distribution of sedimentary microfacies,gravel size,and clay content in fan delta reservoirs,and influences spatial reservoir quality distribution macroscopically.The grain size of the conglomerate is negatively correlated with reservoir quality.The more uniform gravel distribution,the better the sorting,the lower the sorting coefficient,the better the reservoir quality;The clay content in the matrix mainly affects reservoir seepage performance.Compaction is the main diagenetic pore reduction event in the Baikouquan Formation.When the content of carbonate cements is higher than 2.6%,the influence on porosity and permeability increases.The fractal dimension is used to calculate the influence of different types of clay minerals on reservoir quality.Illite and kaolinite have both favorable and detrimental effects on reservoir quality.The fractal dimension value of the best reservoir is generally below 2.5,and the main clay mineral is pore-lining chlorite.The fractal dimension value of the worst reservoir exceeds 2.68,and the main clay mineral is pore-filling chlorite.When the areal porosity of dissolution is greater than 2.2%,the total porosity increases gradually.There is no obvious correlation between the areal porosity of dissolution and permeability.The formation mechanism of high-quality conglomerate reservoirs in the Baikouquan Formation is clarified.The pore evolution of different reservoir architecture is quantitatively reconstructed.The debris flow is mainly composed of medium-grained conglomerate with high clay content and poor sorting.The original porosity of the reservoir is between 29.6% and 31.9%.It mainly underwent rapid and strong compaction during the eodiagenesis,with the porosity decreasing to about 8.9%.During the mesodiagenesis,the reservoir underwent argillaceous and carbonate cementation,with a current porosity of only about 4%.The pore evolution of plain braided channels and debris flow is similar in the eodiagenesis,while some reservoirs experienced late dissolution in the mesodiagenesis,and the current porosity can reach 8%.The original porosity of underwater distributary channel conglomerate reservoirs ranges from 38.8%to 43.5%.Due to its good sorting and the pore retention effect of the chlorite coating,the porosity can be preserved at around 20% until the late eodiagenesis.The selective first oil charge changed the diagenetic evolution process.The compaction and cementation are inhibited by the oil charge.Thus the porosity loss is small,and the current porosity is still as high as about 13%.The reservoir without the first stage of oil charge underwent compaction and cementation in the mesodiagenesis,with a current porosity of about 7.5%.The distribution of excellent reservoirs is determined and their development pattern is established.Through the superposition of lithofacies and diagenetic facies,12 types of petrophysical facies are determined.The reservoir quality classification scheme is determined by clustering of mercury intrusion curves,and the genesis mechanism of excellent reservoirs is determined.A new deep learning porosity prediction method based on prior geological knowledge is proposed,and the average absolute error is reduced from0.83% to 0.48% compared with the conventional deep learning model.Afterward,the planar porosity simulation is done using Bayesian algorithm.The density probability distribution of the porosity simulated by the Bayesian collaborative simulation algorithm is very similar to that of the hard data,as well as the cumulative probability distribution.Based on the classification scheme of reservoir quality,the distribution of high-quality reservoirs is determined by integrating reservoir architecture,interpretation of reservoir type,and porosity prediction result.Consequently,the reservoir development model for different fault blocks is established based on hydrocarbon accumulation mode,reservoir model,and thin section characteristics.