【摘要】 C3相干具有较好的煤层小断层识别能力,但其易受随机噪声的影响,在信噪比较低的情况下容易产生假异常.另一方面,基于S变换的谱分解可以提高地震资料的时频分辨率,并对随机噪声具有一定的压制作用.为提高煤矿采区煤层小断层解释精度,结合谱分解技术与C3相干技术,以提高煤层小断层的解释精度和可靠性.本文分别利用正演剖面和实际三维地震数据,对所提出的方法进行了验证.首先建立了含不同落差的单一煤层模型,利用正演模拟方法获得了其正演剖面和含白噪声剖面.对于这两类剖面,利用S变换获得了其对应于不同频率的谱分解剖面.通过沿煤层反射波提取C3相干属性,获得了不同剖面所对应的C3相干值.对于无噪声地震剖面来说,文中所述方法能明显提高小断层的识别率;对于含白噪声的地震剖面来说,文中所述方法能显著提高所有落差断层的识别率,并基本克服了C3相干算法在低信噪比低情况下容易产生假异常的缺点.将文中所述方法应用到济宁二号煤矿15301工作面的实际三维地震数据中,发现较高频段的谱分解剖面所计算的C3相干同样比常规剖面所计算的C3相干有更好的小断层识别能力.对于煤系地层地质构造较简单的采区来说,利用较高频段谱分解剖面计算的C3相干对落差小于3m的小断层有明显反映.经过井巷工程的实际验证,所解释小断层的可靠性和精度都较高.但当煤系地层地质构造较复杂时,识别落差小于3m小断层的难度较大,还需进一步研究.总之,将谱分解技术和C3相干技术相结合,利用较高频段谱分解剖面计算C3相干,能够明显压制低信噪比所产生的假相干异常,提高煤层小断层的识别精度和可靠性.更多还原

【Abstract】 According to coal seismic data,C3 coherent technique has good detectability of coal bed’s fine faults,but the results is easy affected by random noise.On the other hand,the Spectral Decomposition of S-Transform can improve the resolution of seismic data and suppress random noise.So we combined them together and formed a new technology to make a better interpretation of coal bed’s faults.In order to validate the feasibility of this technology,we utilized both forward modeling sections and actual 3D seismic data.At first,we established a geological model containing a single coal bed,and obtained its forward modeling section.After that,10% random noise was added into this section,and we achieved a new section containing white noise.Then,C3 coherences along coal bed’s reflection of all sections,including seismic sections and their ISO-frequency sections,were extracted along coal bed.By comparing the C3 coherences of forward modeling section and its ISO-frequency section,we found that the interpreted fine faults of coal bed from ISO-frequency section have higher reliability and precision than normal seismic section.Comparatively to seismic section containing white noise,the detected fine faults from its ISO-frequency section with middle-high frequency were not seriously affected by random noise,and basically avoided the false anomalies of C3 coherence.After applied above method to actual 3D seismic data on 15301 workface in Jilin-2 coalmine,we also found that the ISO-frequency section with middle-high frequency has better identifiable ability than normal seismic section in fine faults detecting.That is to say,in relatively simple condition of coal geological structure,fine faults which fault throws are less than 3m are detectable by middle-high ISO-frequency section.By underground validating,the precision and reliability of interpreted fine faults are acceptable and reasonable.On the contrary,when geological conditions of coal-bearing strata are relatively complex,it is very hard to detect these kinds of fine faults.Thus,some further research is necessary.In a sum,the joint technique of spectral decomposition and C3 coherence can improve the detecting precision and reliability of fine faults of coal bed.However,it is still a long way to go to fully identify the fine faults which fault throws are less than 3 m.更多还原