please enter please select

Taking seismic data into the cloud

Project Summary

The UKCS is estimated to have 180,000 sq km of 3D seismic data coverage, amounting to some 35 petabytes worth of seismic data. Subsurface data storage volumes, which are dominated by seismic data, require capital investment in on premise servers. A significant proportion of this data is forced offline to be held on physical tape media due to the sheer volume of it, making it hard for geologists to access and use.

Osokey focused on using modern public cloud infrastructure coupled with its own software and subsurface domain knowledge to enable a serverless, pay as you go approach to seismic data use, visualisation and management. The approach would make seismic data more accessible and remove the periodic capital expenditure of on-premise data servers.

Native format SEGY was moved from inaccessible physical tape archives to the public cloud, with automated ingestion of data using machine learning techniques. The data was then available via a serverless web portal for ease of data access anytime and anywhere. The universal data format was also made available via an Osokey API to Schlumberger’s Petrel application enabling multiple end users to work on a single entity of object stored. SEGY data in a highly performant manner.

Osokey also demonstrated the presentation of additional seismic data types (beyond scope) using its serverless visualisation technology enabling seismic angle stack, 4D seismic, pre-stack data and SEGD field data to be migrated as object stored data into the public cloud.


Industry value:
Given a confirmed data storage cost saving from this project of $60,000/100TB per year, it is estimated that moving the UKCS seismic data storage to the public cloud could collectively save the industry $21 million per year (£16.8 million).

Key results / Lessons learned:
This project delivered seismic in the cloud technology, storing and archiving native SEGY and SEGD seismic files in a cloud agnostic manner.

This project demonstrated that cost savings in excess of 70% could be achieved in the storage of seismic data. Cloud storage costs were quantified during the project. Traditional storage of a benchmark volume of 10 terabytes can cost $750 per month in comparison to cloud prices of less than $250 per month. This was further verified with proprietary Chevron cost estimations.

It also showed that having seismic data available on the cloud enables rapid machine learning interpretations of subsurface regional geology. By using the cloud’s scalable cloud computation, approximately 600,000 files were analysed in less than 30 minutes and identified 10% of the files were unnecessary duplicates offering a further storage cost saving for a miniscule cost of ~$15 of cloud compute.

The automated data ingestion capability removed a previous bottleneck to the adoption of cloud technology for seismic data. Additional benefits included the ability to remove data duplication from workflows.

The Oil & Gas Authority have since named Osokey as their collaborative partner to develop the National Data Repository (NDR) platform.  [Osokey to further develop UK data repository | Offshore (offshore-mag.com)]


Related

Let's work together

"Required" indicates required fields

My name isRequired
My name is
This field is for validation purposes and should be left unchanged.