Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:

insideHPC Guide to HPC/AI for Energy – Part 3

In this technology guide sponsored by our friends over at Dell Technologies, we take a deep dive into how the team of Dell Technologies and AMD is working to provide solutions for a wide array of needs for more strategic cultivation of oil and gas energy reserves. We’ll start with a series of compelling use-case examples, and then introduce a number of important pain-points solved with HPC and AI. We’ll continue with some specific solutions for the energy industry by Dell and AMD. Then we’ll take a look at a case study examining how geophysical services and equipment company CGG successfully deployed HPC technology for competitive advantage. Finally, we’ll leave you with a short-list of valuable resources available from Dell to help guide you along the path with HPC and AI.

This technology guide, insideHPC Guide to HPC/AI for Energy, explores how the dynamic confluence of HPC and AI is now considered essential to any organization involved in energy exploration and the process of bringing the resulting energy resources to market.

Energy Industry Pain-Points Solved by HPC/AI

The energy industry has a number of critical pain-points that may be solved through use of HPC and AI  technologies. In this section we’ll highlight a number of these pain-points and illustrate ways that HPC and AI  can overcome these problematic roadblocks.

At a high level, energy companies experience increasing difficulty finding and extracting energy resources, so the risks of making erroneous decisions are greater than ever. As exploration goes deeper and into harsher  environments, the risk and costs of miscalculating drill sites increases.

The costs of finding and then bringing discoveries into production are higher than in the past. In addition,  shrinking margins make IT technology efficiency a critical success factor for industry players. Geologists need  timely and accurate information to make the best decisions. Specifically, the efficiencies represented by HPC  and AI are recognized to provide a significant return on investment (ROI) for companies deploying these technologies.

Energy exploration

AI provides energy exploration companies the assurance to isolate drilling locations, resulting in cost savings.  Deploying a combination of HPC and AI algorithms, optimal drilling locations can be determined  with the awareness that the location along with the depth and direction of a well will supply the maximum  amount of energy at a minimized cost.

A serious pain-point is the long path from exploration to drilling–the initial surveying to transporting oil to  where it is demanded can take years of planning, exploration and lastly drilling.

Once the drilling has begun, the process requires additional resources for observation and analysis. The cost  of a new well may be in the 100s of millions of dollars, so it is critical that accurate decisions be made early in  the process to avoid costly planning changes down the line. This is where HPC and AI can play an important role.

The amount of data produced by a seismic survey can be significant, often in the terabyte to petabyte range.  This data, produced in the field, must then be brought into a seismic processing system to determine the  location of the reserves and the most optimum methods to extract the oil or natural gas. Significant hardware  resources are required–servers, storage and networking in order to properly analyze the raw data.

Geophysics simulation

Well-conceived geophysics simulation is another important pain point encountered by energy companies.  The process has been used for many years to locate underground deposits. The amount of data that must be  analyzed is growing in order to make more precise decisions for locating the most likely drilling locations.  Increasingly, GPU-powered deep learning is being used to solve the compute requirements for simulations:

  • Reduces costs by enabling more comprehensive, accurate, and detailed visualizations and surveys
  • Increases overall productivity to further improve data-driven decision-making capabilities
  • Handles massive data-sets and time-critical computing processes with ease
  • Compatible with top applications used for geophysical research and exploration.

The area of simulation is a prime example of where deep learning using artificial neural networks (ANNs) can  play a significant role in enhanced understanding of what exists in subsurface locations. ANNs, utilizing large  numbers of computing elements such as those found in modern GPUs, take complex data and compare it to  earlier patterns in order to lead to making more accurate decisions. The need for simulation based on the  current understanding of nature remains as strong as ever, but can now be combined with AI algorithms that  have been trained to recognize certain data models. Incorporating previously learned knowledge  automatically into the discovery process leads to a decrease in exploration costs.

Pipeline integrity

Pipeline integrity is another important pain point. With miles of pipelines contained within and external to a  drilling rig, an early perception of possible failure points can result in significant cost and environmental  savings. Using a variety of strategic data collection methods, AI-based systems can employ alert operators to  identify possible upcoming failures.

Edge computing

While simulations that take advantage of thousands of processors can deal with data that has already been  pre-processed, another industry pain point is an increasing need to process the tremendous amounts of data  produced before the simulation even begins. So-called “streaming data” must be ingested and turned  into meaningful data that can be used. Edge computing is technology that can help alleviate this data deluge.

As part of the overall exploration workflow that then leads into production, a number of areas exist in which  computing at the edge greatly benefits the process and can lead to reduced time to insights and decisions. For example, if the terabytes of data generated at the source can be filtered locally, less data must be sent  over unreliable or non-existent networks that may exist in areas far away from reliable communication  devices. The advantage is not just in filtering data at the edge of the network, but also in relying on previous  patterns and known data sets that are available. Sophisticated AI algorithms can be trained to recognize  when data is reliable or when it could be discarded without human interaction.

Drilling rigs are very complex to build and operate. Thousands of sensors are in play and must be monitored  and corrective actions must be taken depending on the data collected. Through deep learning and edge  computing, an accurate understanding of the operation of the drilling platform can be attained–problems  can be identified and resolved in a time expedient fashion.

Over the next few weeks we will explore how with the accelerated use of HPC and AI in the energy industry,  there are strategic options for laying a directed path for these technologies to become a strong competitive advantage. Dell Technologies and AMD is a team that can help you travel down this path:

Download the complete insideHPC Guide to HPC/AI for Energy, courtesy of Dell.

Leave a Comment


Resource Links: