A Cloud Approach to Processing Distributed Vibration Measurements

Distributed vibration sensing (DVS), over a typical well length of 3.5 km, provides 410 MB/s of data, which is a lot to process and handle. Ideally, these data are also made available for interpretation and insight in real-time.

The very large data volume involved requires the data, algorithms, and compute resources to unite efficiently. The computing resources may consist of an offshore platform's on-premises compute; or there may be partial compute at well-site followed by uploading to the cloud, with further processing applied there. Physical logistics, such as when to retrieve physical hard-drives to shore for general archiving, must also addressed in a DVS system.

We outline a cloud-based reprocessing system for DVS. This system is fully elastic and so can re-compute the entire data-volume on-demand. The approach also enables new processing techniques and we present examples of the very low frequency (VLF) response and the trade-off between signal distortion and spatial resolution.