Extraction of Reflections from Tube Waves

Meng (1997) used the McConnell crosswell data to show that transmitted arrival traveltimes can be extracted from common receiver gathers (CRG) where the effective source is a tube wave generator. In this report I use the McElroy crosswell data to show that reflection events can be extracted from CRG’s where the effective source is a tube wave generator. This result opens the possibility that more economical crosswell seismic survey might be possible using surface seismic sources and tube wave generators in the source well. The remaining question is whether or not the tube wave generator is strong enough to generate high signal to noise ratio reflections. INTRODUCTION Meng (1997) proposed the possibility that a crosswell survey can be performed by only using a surface seismic source and tube wave generators (i.e., obstructions) in the source well. As an example, the tube wave response in the source well was computed for Noranda’s McConnell crosshole data and the corresponding transmitted direct waves were clearly imaged. However, the reflection waves were too weak for that particular tube wave secondary source as well as for the conventional seismic source. This leaves open a question: Is the tube wave generator strong enough to generate reflection signals which are needed for crosswell migration? In this report I process the McElroy crosswell data to answer this question. Results show that the tube wave responses extracted from the McElroy crosswell data contain useful direct and reflected P and S waves which can be used for crosswell imaging. EXTRACTING TUBE WAVE RESPONSE FROM CONVENTIONAL CROSSWELL DATA The processing needed for extracting the tube wave responses is as follows: 1. Supposing there exists an impedance discontinuity at depth Z0 in the source well, the tube wave will be generated at this obstruction using a conventional downhole source