地质学
中国
长江
地球化学
采矿工程
古生物学
考古
地理
作者
Brian Horsfield,Caineng Zou,Jian Li,Shengyu Yang,Nicolaj Mahlstedt,David Misch,Doris Groß,Wei Ma,Yifeng Wang,Jingqiang Tan
出处
期刊:AAPG Bulletin
[American Association of Petroleum Geologists]
日期:2021-05-01
卷期号:105 (5): 945-985
被引量:7
摘要
China has been said to have the largest putative shale gas resources in the world. The highest potential occurs in the Sichuan Basin, with the overmature Qiongzhusi (Cambrian) and Longmaxi (Silurian) Formations as prime exploration targets. Here, the likelihood of late gas formation is examined using less mature equivalents from the Georgina Basin (Australia) and the Baltic Basin (Lithuania). We consider the respective functions of kerogen and polar bitumen in gas generation with reference to the Eagle Ford, Yanchang, Niobrara, and Vaca Muerta Formations. Both of the lower Paleozoic shales are bitumen poor in a geochemical sense, this being in stark contrast to the Mesozoic shales, which are bitumen rich. Kerogen is, therefore, the major gas precursor in the Cambrian and Silurian of the Sichuan Basin. Graptolites and solid bitumen are petrographically dominant. The solid bitumen exhibits flow structures and is deduced to be highly polar because it is insoluble in dichloromethane. Secondary cracking kinetics determined for the Arthur Creek are closely similar to source rocks containing predominantly paraffinic oil. Late gas generation from very stable refractory kerogen structures via alpha-cleavage reactions at maturities above 2% equivalent vitrinite reflectance (Ro) was verified, and importantly, the upper ceiling for late gas generation in Paleozoic shales of the Sichuan Basin is set at 3% Ro. As far as the Qiongzhusi shale is concerned, raising the prospective acreage to a 3% Ro limit brings an additional contribution of 40 mg HC/g total organic carbon of late gas charge. The same is true for the extensive fairway of the Longmaxi shale along the western flank of the basin, close to the subcropping erosional edge.
科研通智能强力驱动
Strongly Powered by AbleSci AI