Increasing volumes of seismic data being analyzed by geoscientists, increased stress on the need to reduce error margins in terms of prospect generation, the need to lower the probability of dry holes and an increased need to speed up the time to first oil ensures that High Performance Computing solutions are a key component in E&P. However, fluctuating costs, the need for de-localization, operating norms and increasing levels of uncertainties have placed restrictions on the amount of HPC solutions developed in-house. To help with this, Computational Research Laboratories Ltd. (CRL) brings HPC services and solutions for Oil & Gas enterprises. With services ranging from leveraging the world’s fastest commercially available supercomputer for data processing and reservoir simulation activities to development of applications for Reverse Time Migration and Wave Form Inversion, CRL provides a suite of services to partner with your enterprise at all levels that require HPC solutions.

Challenges :

1. The volumes of seismic data being analyzed by geoscientists have grown tremendously. 3-D (three -dimensional) volumes are used routinely for structural, stratigraphic and pore fluid detection. Petroleum geophysicists have had to compromise, often accessing only a portion of the full information that could be gathered from a given dataset if high-performance computing (HPC) were more easily and affordably accessible. Their algorithms and iterative procedures continue to evolve and demand significant memory allotments and processing time. The waiting in line for data processing, wherever that processing is done, delays -decision-making and thwarts people’s performance.

2. With the introduction of operating norms, it becomes an extremely tedious and cumbersome process to use cross-continent HPC systems when servicing specific geographies. Access to a local HPC system along with the domain expertise in data processing and reservoir simulation provides the added boost to smoothen the execution when servicing E&P companies in restricted geographies.

3. The augmentation of existing HPC systems or installation of new facilities in the face of uncertainties, fluctuating prices and volatile demands has proved to be restrictive for Research & Development activities. Operating activities like seismic processing, reservoir modelling, dynamic time- based fluid modelling and closed loop analytical processes increasingly absorb the available computing capacities.

Our Solution:

At CRL we bring to the table a complete ecosystem of a production ready state-of-the-art HPC infrastructure, domain expertise in upstream E&P activities and in-house HPC expertise to support E&P needs in the compute intensive areas of seismic data processing and reservoir modeling.
We provide the following components to support HPC solutions required in E&P activities:

Infrastructure Provisioning : Ready access to the world’s fastest commercially available supercomputer ‘Eka’. With 1800 production ready compute nodes, 20 Gbps high speed interconnect and high throughput scalable storage providing a peak performance of 172 TeraFlops; your enterprise has immediate access to a powerful HPC infrastructure without the need for capital investments or long term lease options.

Application Services : The application domain team at CRL comprises of experts with wide experience in E&P companies and service providers. CRL provides the expertise required to port and commission various data processing and reservoir simulation applications on a High Performance Computing infrastructure. User workflows can be optimized and customized based on varying requirements within short turnaround times.

Remote Access : CRL provides secure remote access mechanisms for submissions, monitoring and intermediate data transfers to review the execution results.

Parallel Programming : The HPC expertise at CRL extends to parallel programming on using MPI, OpenMPI/pThread based or even hybrid methodologies. Parallelization using libraries like HP MPI, Intel MPI, OpenMPI and Mvapich, among others, is done at CRL to successfully improve scalability and performance of existing application codes.