Why keep pre-stack seismics near-line? (January 1998)

We have been fretting of late as to why somecompanies keep their field seismic data on or near-line and others dispatch it to shelvingor even off-site storage. We asked Calgary-based Kelman Archives, a spin-off of KelmanSeismic Processing Inc. why this was one of their key offerings. Neil Baker Kelman's VPMarketing told us the following.

We here at Kelman have several reasons for offering pre-stack seismic data on demand. The seismic processing shops that we serve actually prefer to process their data from disk. This obviously reduces the time required to read the data into their systems for demultiplex and speeds the turnaround cycle for all our common clients. We have supplied processing companies with the software required to read in field data from disk The data is received in its' recorded format so that there are no changes required for the receiving company in their demultiplex programs. Tape mounts are zero and data is read in at disk speed.

Our oil company clients using AVO and other pre-stack interpretive tools have found these to be computationally cumbersome in the past due to high disk space requirements and long processing times. To turn these into viable pre-stack processes we have created an interactive demultiplex module which allows our users to retrieve pre-stack data from the robot and demultiplex on the fly. This can be performed on either single or multiple records.

Elsewhere, clients may be focused on seismic field program design. Here legacy field data can be analyzed from the workstation so that field parameter selection can be optimized. During seismic acquisition, clients with in-house processing capabilities can now easily view, test and analyze the data as part of normal QC procedures - even in isolated areas.

In Canada trading and brokering data is a large component of geophysical activity. With pre-stack data on-line, together with other imaged data such as observers logs, survey, and other supporting documentation, purchased data can be accessed and delivered instantly.

Our system is also an on-line disaster recovery tool that has two robots for "mirrored data". With the current state of most libraries containing only one copy of data and the known problems of stiction causing data loss this system allows us to recover and protect this asset for the future. We believe that field data is for using, not just for storing. Companies can only exploit their data to the full - especially in the field of high-tech pre-stack processing - if it is accessible.