煤层含气量的测试、模拟与预测研究进展

Review of research on testing,simulation and prediction of coalbed methane content

煤层含气量是煤层气资源勘探和开发的关键参数,我国煤层气(瓦斯)含量直接测试是通过地面或井下钻孔钻取煤心在大气压条件下解吸,一方面取心和装罐过程中煤样中部分水溶气、游离气,甚至发生解吸的极少部分吸附气被逸散;另一方面在储层压力直接降为大气压的解吸过程中部分气体被“气锁”不能完全解吸,均导致煤层含气量实测值偏低.本文对比分析了我国煤层含气量测试方法、测试标准的演化历程,系统评述了校正系数、气体相态、带压逐级降压解吸、地史演化、煤层气井产气量历史匹配等物理模拟和数值模拟煤层含气量的可靠性及含气梯度法、煤级-压力-等温吸附曲线法、地球物理解释技术、人工智能技术等预测方法的适应性,认为增加样品重量至2kg、减少损失时间至小于5min、增长解吸时间至24h、稳定在煤层原位温度、保持样品压力能提高煤层含气量直接测试的精度,密封保压取心能减少损失气估算的误差;测试与模拟相结合的方法不仅可以缩短测试时间至1d,而且还能获得煤层全部水溶气、游离气含量;钻孔密封保压取心后在储层温度下带压逐级降压解吸所测试的煤层含气量能够反映原位煤层的真实含气量;三相态含气量、带压逐级降压解吸、校正系数和煤层气井产气量历史匹配可以获得接近原位煤层的含气量;依据地质条件合理选择煤层含气量预测方法,包含煤岩、煤质、围岩封闭性、埋深和地球物理响应值的人工智能技术预测的煤层含气量可信度高.研究结果表明:沁水煤田晋城矿区无烟煤储层煤层气资源勘探实测煤层含气量是煤炭资源勘探实测煤层含气量的1.32~1.44倍,煤层气产量历史匹配与带压逐级降压解吸获得的煤层含气量是煤层气资源勘探实测煤层含气量的1.17~1.46倍.

The coalbed methane(CBM)content is a key parameter for the exploration and development of CBM.In China,the direct measurement of CBM content or coal gas content is based on the desorbing of coal sample collected from surface or underground wellbore under atmospheric pressure.On the one hand,part of water-soluble gas,free gas,and even a very small proportion of adsorbed gas(desorbing from coal sample)may escape during coring and canning.On the other hand,when the reservoir pressure decreases directly to the atmospheric pressure level,part of methane cannot be desorbed completely due to gas-locked.The above two reasons can result in a low measured gas content compared to the real gas content value.This paper compared and analyzed the evolution process of the CBM content measurement methods and standards,and systematically reviewed the reliability of physical simulation or numerical simulation technologies for gas content assessment,including the correction coefficient,phase of gas,decompression and desorption under successively decreasing pressure,evolution of geological history,and gas production history matching of CBM well.The adaptability of prediction methods was also analyzed,such as gas gradient method,coal rank-pressure-isothermal adsorption curve method,geophysical interpretation methods,and artificial intelligence methods.The results showed that the accuracy of gas content testing can be improved by increasing the sample weight to 2kg,reducing the loss time to less than 5min,increasing the desorption time to 24h,stabilizing the in-place temperature of coal seam,and maintaining the sample pressure as well.Maintaining the sample pressure while coring can reduce the error of lost gas estimation.Combining the testing methods and simulation can not only shorten the testing period to 1d,but also obtain the contents of total water-soluble gas and free gas.After drilling pressure-maintaining and coring,the gas content of coal measured by desorption under successively decreasing pressure at the reservoir temperature can well reflect the original gas content.Three-phase gas content,step-by-step decreasing pressure desorption,correction coefficient,and historical matching of gas production from coalbed methane wells can obtain a gas content value close to the original one.It is necessary to select appropriate CBM content prediction technologies according to geological conditions.The gas content predicted by artificial intelligence technology,including coal petrology and quality,surrounding rock sealing,buried depth,and geophysical response values,is believed to be more reliable.The measured anthracite gas content during CBM resource exploration in the Qinshui Coalfield,Jincheng Mining Area is 1.32—1.44times than that obtained by coal exploration.And the gas content obtained by historical matching of CBM production and step-by-step decreasing pressure desorption is 1.17—1.46times than that obtained by CBM resource exploration.