Wells CXD1 and CX2 have uncovered high-concentration potassium-and lithium-containing brines and substantial layers of halite-type polyhalite potash deposits within the 4th and 5th members of the Triassic Jialingjiang...Wells CXD1 and CX2 have uncovered high-concentration potassium-and lithium-containing brines and substantial layers of halite-type polyhalite potash deposits within the 4th and 5th members of the Triassic Jialingjiang Formation and the 1st Member of Leikoupo Formation(Jia 4 Member,Jia 5 Member,and Lei 1 Member)in the Puguang area,Sichuan Basin.These discoveries mark significant breakthroughs in the exploration of deep marine potassium and lithium resources within the Sichuan Basin.Utilizing the concept of“gas-potassium-lithium integrated exploration”and incorporating drilling,logging,seismic,and geochemical data,we have investigated the geological and enrichment conditions,as well as the metallogenic model of potassium-rich and lithium-rich brines and halite-type polyhalite.First,the sedimentary systems of gypsum-dolomite flats,salt lakes and evaporated flats were developed in Jia 4 Member,Jia 5 Member,and the 1st member of Leikoupo Formation(Lei 1 Member)in northeastern Sichuan Basin,forming three large-scale salt-gathering and potassium formation centers in Puguang,Tongnanba and Yuanba,and developing reservoirs with potassium-rich and lithium-rich brines,which are favorable for the deposition of potassium and lithium resources in both solid or liquid phases.Second,the soluble halite-type polyhalite has a large thickness and wide distribution,and the reservoir brine has a high content of K+and Li+.A solid-liquid superimposed“three-story structure”(with the lower thin-layer of brine reservoir in lower part of Jia 4 Member and Jia 5 Member,middle layer of halite-type polyhalite potash depositS,upper layer of potassium-rich and lithium-rich brine reservoir in Lei 1 Member)is formed.Third,the ternary enrichment and mineralization patterns for potassium and lithium resources were determined.Vertical superposition of polyhalite and green bean rocks is the mineral material basis of potassium-lithium resources featuring“dual-source replenishment and proximal-source release”,with primary seawater and gypsum dehydration as the main sources of deep brines,while multi-stage tectonic modification is the key to the enrichment of halite-type polyhalite and potassiumlithium brines.Fourth,the ore-forming process has gone through four stages:salt-gathering and potassium-lithium accumulation period,initial water-rock reaction period,transformation and aggregation period,and enrichment and finalization period.During this process,the halite-type polyhalite layer in Jia 4 Member and Jia 5 Member is the main target for potassium solution mining,while the brine layer in Lei 1 Member is the focus of comprehensive potassium-lithium exploration and development.展开更多
Pore structures in shales are a main factor affecting the storage capacity and production performance of shale gas reservoirs.Taking Longmaxi Shales in the Jiaoshiba area of the Sichuan Basin as a study object,we syst...Pore structures in shales are a main factor affecting the storage capacity and production performance of shale gas reservoirs.Taking Longmaxi Shales in the Jiaoshiba area of the Sichuan Basin as a study object,we systematically study the microscopic pore structures of shales by using Argon-ion polishing Scanning Electron Microscope(SEM),high-pressure mercury injection and low-temperature nitrogen adsorption and desorption experiments.The study results show that:the Longmaxi Shale in this area are dominated by nano-scale pores which can be classified into organic pores,inorganic pores(intergranular pores,intragranular pores,inter-crystalline pores and dissolution pores),microfractures(intragranular structure fractures,interlayer sliding fractures,diagenetic shrinkage joints and abnormal-pressure fractures from organic evolution),among which organic pores and clay mineral pores are predominant and organic pores are the most common;a TOC value shows an obvious positive correlation with the content of organic pores,which account for up to 50%in the lower-quality shales with a TOC of over 2%where they are most developed;microscopic pore structures are very complex and open,with pores being mainly in cylinder shape with two ends open,or in parallel tabular shape with four sides open and 2–30 nm in diameter,being mostly medium pores.On this basis,factors affecting the micropore structures of shales in this area are studied.It is concluded that organic matter abundance and thermal maturity are the major factors controlling the microscopic pore structures of shales,while the effects of clay mineral content are relatively insignificant.展开更多
目的基于机器学习算法,构建太原市乙型肝炎表面抗原(hepatitis B surface antigen,HBsAg)阳性母亲婴儿乙型肝炎疫苗无/弱应答的风险预测模型,为此类高风险婴儿的早期识别和预防提供参考。方法选取2019年11月—2023年10月在太原市第三人...目的基于机器学习算法,构建太原市乙型肝炎表面抗原(hepatitis B surface antigen,HBsAg)阳性母亲婴儿乙型肝炎疫苗无/弱应答的风险预测模型,为此类高风险婴儿的早期识别和预防提供参考。方法选取2019年11月—2023年10月在太原市第三人民医院妇产科分娩的HBsAg阳性产妇及其分娩的婴儿作为研究对象,采用问卷调查和病历查阅方式收集数据。采用多因素logistic回归分析模型筛选危险因素,采用5折交叉验证划分数据集,采用合成少数样本过采样方法对训练集数据进行重采样处理后,分别构建logistic回归、极端梯度提升(extreme gradient boosting,XGBoost)、决策树、随机森林模型,并采用受试者工作特性曲线下面积(area under the curve,AUC)评价模型预测性能。结果共收集到HBsAg阳性母婴253对,其中无/弱应答率为10.28%(26/253)。Logistic回归分析模型结果显示,母亲乙型肝炎病毒DNA载量、新生儿髓样分化因子88蛋白百分比、磷酸化核因子κB蛋白百分比、浆细胞百分比是婴儿乙型肝炎疫苗无/弱应答的影响因素。将以上因素纳入预测模型,在验证集中各模型AUC均>0.800,其中以XGBoost算法构建的预测模型效能最佳(AUC=0.919)。结论基于机器学习算法,尤其是XGBoost算法,建立的太原市HBsAg阳性母亲婴儿乙型肝炎疫苗无/弱应答风险预测模型具有较好的效能,有助于预测婴儿乙型肝炎疫苗无/弱应答发生。展开更多
基金Supported by the National Research and Development Program(2017YFC0602804)Geological Bureau Program of Sichuan Province(SCDZ-KJXM202403).
文摘Wells CXD1 and CX2 have uncovered high-concentration potassium-and lithium-containing brines and substantial layers of halite-type polyhalite potash deposits within the 4th and 5th members of the Triassic Jialingjiang Formation and the 1st Member of Leikoupo Formation(Jia 4 Member,Jia 5 Member,and Lei 1 Member)in the Puguang area,Sichuan Basin.These discoveries mark significant breakthroughs in the exploration of deep marine potassium and lithium resources within the Sichuan Basin.Utilizing the concept of“gas-potassium-lithium integrated exploration”and incorporating drilling,logging,seismic,and geochemical data,we have investigated the geological and enrichment conditions,as well as the metallogenic model of potassium-rich and lithium-rich brines and halite-type polyhalite.First,the sedimentary systems of gypsum-dolomite flats,salt lakes and evaporated flats were developed in Jia 4 Member,Jia 5 Member,and the 1st member of Leikoupo Formation(Lei 1 Member)in northeastern Sichuan Basin,forming three large-scale salt-gathering and potassium formation centers in Puguang,Tongnanba and Yuanba,and developing reservoirs with potassium-rich and lithium-rich brines,which are favorable for the deposition of potassium and lithium resources in both solid or liquid phases.Second,the soluble halite-type polyhalite has a large thickness and wide distribution,and the reservoir brine has a high content of K+and Li+.A solid-liquid superimposed“three-story structure”(with the lower thin-layer of brine reservoir in lower part of Jia 4 Member and Jia 5 Member,middle layer of halite-type polyhalite potash depositS,upper layer of potassium-rich and lithium-rich brine reservoir in Lei 1 Member)is formed.Third,the ternary enrichment and mineralization patterns for potassium and lithium resources were determined.Vertical superposition of polyhalite and green bean rocks is the mineral material basis of potassium-lithium resources featuring“dual-source replenishment and proximal-source release”,with primary seawater and gypsum dehydration as the main sources of deep brines,while multi-stage tectonic modification is the key to the enrichment of halite-type polyhalite and potassiumlithium brines.Fourth,the ore-forming process has gone through four stages:salt-gathering and potassium-lithium accumulation period,initial water-rock reaction period,transformation and aggregation period,and enrichment and finalization period.During this process,the halite-type polyhalite layer in Jia 4 Member and Jia 5 Member is the main target for potassium solution mining,while the brine layer in Lei 1 Member is the focus of comprehensive potassium-lithium exploration and development.
文摘Pore structures in shales are a main factor affecting the storage capacity and production performance of shale gas reservoirs.Taking Longmaxi Shales in the Jiaoshiba area of the Sichuan Basin as a study object,we systematically study the microscopic pore structures of shales by using Argon-ion polishing Scanning Electron Microscope(SEM),high-pressure mercury injection and low-temperature nitrogen adsorption and desorption experiments.The study results show that:the Longmaxi Shale in this area are dominated by nano-scale pores which can be classified into organic pores,inorganic pores(intergranular pores,intragranular pores,inter-crystalline pores and dissolution pores),microfractures(intragranular structure fractures,interlayer sliding fractures,diagenetic shrinkage joints and abnormal-pressure fractures from organic evolution),among which organic pores and clay mineral pores are predominant and organic pores are the most common;a TOC value shows an obvious positive correlation with the content of organic pores,which account for up to 50%in the lower-quality shales with a TOC of over 2%where they are most developed;microscopic pore structures are very complex and open,with pores being mainly in cylinder shape with two ends open,or in parallel tabular shape with four sides open and 2–30 nm in diameter,being mostly medium pores.On this basis,factors affecting the micropore structures of shales in this area are studied.It is concluded that organic matter abundance and thermal maturity are the major factors controlling the microscopic pore structures of shales,while the effects of clay mineral content are relatively insignificant.
文摘目的基于机器学习算法,构建太原市乙型肝炎表面抗原(hepatitis B surface antigen,HBsAg)阳性母亲婴儿乙型肝炎疫苗无/弱应答的风险预测模型,为此类高风险婴儿的早期识别和预防提供参考。方法选取2019年11月—2023年10月在太原市第三人民医院妇产科分娩的HBsAg阳性产妇及其分娩的婴儿作为研究对象,采用问卷调查和病历查阅方式收集数据。采用多因素logistic回归分析模型筛选危险因素,采用5折交叉验证划分数据集,采用合成少数样本过采样方法对训练集数据进行重采样处理后,分别构建logistic回归、极端梯度提升(extreme gradient boosting,XGBoost)、决策树、随机森林模型,并采用受试者工作特性曲线下面积(area under the curve,AUC)评价模型预测性能。结果共收集到HBsAg阳性母婴253对,其中无/弱应答率为10.28%(26/253)。Logistic回归分析模型结果显示,母亲乙型肝炎病毒DNA载量、新生儿髓样分化因子88蛋白百分比、磷酸化核因子κB蛋白百分比、浆细胞百分比是婴儿乙型肝炎疫苗无/弱应答的影响因素。将以上因素纳入预测模型,在验证集中各模型AUC均>0.800,其中以XGBoost算法构建的预测模型效能最佳(AUC=0.919)。结论基于机器学习算法,尤其是XGBoost算法,建立的太原市HBsAg阳性母亲婴儿乙型肝炎疫苗无/弱应答风险预测模型具有较好的效能,有助于预测婴儿乙型肝炎疫苗无/弱应答发生。
