| Authors: |
Byoungjoon YOON, Kwon Gyu PARK, Changhyun LEE, Kichang YANG, Yoonho SONG, Tae Jong LEE, Inhwa PARK |
| Keywords: |
seismic while drilling, pre-processing, top-drive pilot, geothermal well, Pohang, Korea, |
| Conference: |
World Geothermal Congress |
Session: |
Geophysics |
| Year: |
2015 |
Language: |
English |
| Abstract: |
Seismic While Drilling (SWD) surveys that use drill-bit vibration energy as a source were conducted to acquire velocity information in the vicinity of geothermal well in Pohang, Korea in 2012 and 2013. The cross-correlation between a pilot signal and geophone signals is a primary processing step for SWD data. Hence sensitivity of the pilot sensor becomes a crucial factor for SWD processing. Generally, pilot sensors are located on the top of the drill-pipe or in the bottom hole assembly to measure reference drill-bit vibration signals. However, in case of our surveys, two pilot sensors were deployed both on the Blowout Preventer (BOP) and top drive due to the safety issue and time-pressing drilling schedule. As a consequence, Signal to Noise Ratio (SNR) of pilot signals and their simple cross-correlation results were not satisfactory input data for further SWD processing. Accordingly, we focused on exploiting several pre-processing techniques to improve SNR of pilot signals and cross-correlation results. Comparison analysis for pilot signals according to drilling conditions shows that the frequency range from 40 to 140 Hz are reinforced for top-drive pilot signals during drilling operation. This enhancement of the specific frequency band is also observed similarly in the amplitude spectrum of the data observed at a borehole accelerometer located near the drill well. According to these observations, top-drive pilot signals whose 40 to 140 Hz frequency bands are relatively strong were selected for further SWD processing. Because of extremely noisy environments (e.g. electronic noise, mud noise, engine noise, and rig noise), harmonic noise subtraction and adaptive noise subtraction were inevitably needed to refine pilot signals and geophone signals before cross-correlation. Pilot de-convolution and selective stack were also applied to cross-correlograms. As a result of these pre-processing works, the SNR of the cross-correlogram was improved for further SWD processing. |
| Download: |
|
File Size: |
3406 KB |