井下地震计方位角检测技术应用研究

本文利用波形相关分析法对井下地震计方位角检测精度及一致性进行分析研究.研究中在地表布设地震计进行不同距离、不同频带多套地震计方位角检测对比测试,并在四川泸州、宁夏灵武、陶乐、河南安阳、清丰、陕西定边等6个分别安装井下甚宽频带、宽频带、短周期地震计的台站,进行地表同台址不同频带多套地震计与井下地震计方位角检测及一致性对比测试.根据全球噪声模型1~10s之间存在明显的噪声峰值的特点,对测试数据进行0.2~0.3Hz带通滤波和仿真处理,通过分析得出,对于不同地震计组合,测试地震计方位角检测结果一致性较好,根据地表不同距离、不同频带地震计组合的检测精度和地脉动记录的特点,认为井下地震计方位角检测精度优于4°.

Research on the application of borehole seismometer azimuth detection technology

Borehole seismic observation can effectively reduce the impact of ground noise, and improve the observation accuracy of micro seismic events. At present, there are more than 180 borehole seismic stations in China, covering observations from very broad band to short period. Orienting equipment such as gyroscope is generally used in the installation of a borehole seismometer to determine its azimuth.Due to the difference between borehole and surface observation, it is difficult to verify the actual horizontal azimuth of a borehole seismometer directly using a gyroscope or other orienting equipment.A correlation analysis method was used in the azimuth detection of a borehole seismometer. To confirm its reliability and consistency, we first made comparison test of a set of seismometers on the surface with different frequency bands. And then, another comparison test was made on a set of seismometers with different distances apart from each other to verify the result accuracy and correlation coefficient variation trend on different bands and apart distances. We assume that it should be true also in the case of correlation analysis between borehole and surface seismometers.Finally, multiple sets of surface and borehole seismometers comparison tests were made, and their result reliability and consistency are discussed also.Test results at 3 ground surface experiment sites ZOZT, XAN and QLIT, show that measuring cases with the same device distance but different frequency bands may yield azimuth and correlation coefficient with good consistency.As distance between compared devices increased, the bias of estimated azimuth relative to the real one increased, and the correlation coefficient decreased.For the experiment with device distance of 220 m, the root mean square error(RMSE) of estimated azimuths to the true case is less than 1 degree, the averaged correlation coefficient is about 0.97.For the experiment with device distance of 2514 m, the RMSE of estimated azimuths is less than 3.4 degree, and the correlation coefficients vary from 0.84 to 0.80. At 6 borehole stations LWU,TLE, AY, QF, DBT, LZH, multiple surface seismometers of different frequency band with gyroscope direction aligned were mounted near each wellhead. Analyses for different band devices show their results on relative azimuth and correlation coefficient in good agreement. At 5 out of 6 stations as LWU, TLE, AY, QF and DBT, their estimated azimuths deviate greatly from azimuths given by directional equipment used in their installation. It reveals that there are some defects in the orienting method used in the installation of borehole seismometer, and horizontal azimuth error of borehole seismometer exists commonly. At station LZH, the borehole seismometer was installed without directional device orientation, its seismometer azimuth was given during installation by Scream software of Güralp company via estimation with a synchronized and oriented surface seismograph system. Our work at LZH yields a result consistent with the given azimuth. Our works show that correlation analyses between ground surface seismometers with different distances and frequencies yield results consistent with those of analyses between surface and borehole devices. So, the correlation analysis method may be used to estimate the horizontal azimuth of a borehole seismometer with frequency band from very broad band to short period. The correlation analysis is done on microtremor in the frequency band 0.2~0.3 Hz, which has a wave length that is far greater than the depth of borehole in study and is the common dominant microtremor records in both surface record and borehole record. Hence, the accuracy of estimated azimuth of a borehole seismometer may be inferred from the results of experiments on the ground surface occupied seismometers with different apart distances. The results of surface experiments show that the accuracy of estimated azimuth gets lower and correlation coefficient gets smaller as the seismometers apart distance increased. For the instance case of 920 m, 2249 m and 2514 m surface seismometer apart distance, the errors of estimated azimuths are less than 4°. Therefore, considering the impact of various factors on our experiments, we believe that the azimuth estimation accuracy of a borehole seismometer with depth less than 400 m should be better than 4°.