Oil and gas companies have identified analytics and intelligence needs that are not met (or not easily met) by existing manufacturing management, quality, or process control. Investments in process control, process management, and data collection applications over the past 20 years have resulted in a mountain of data, much of which goes unused because very complex integration environments impede the introduction of new applications and functionality. These are the needs that require enterprise manufacturing intelligence (EMI), which delivers:

  • Aggregation: making data available from many sources, most often databases;
  • Contextualization: providing a structure or model for the data that will help users find what they need;
  • Analysis: enabling users to analyze data across sources and especially across production sites;
  • Visualization: providing tools to create visual summaries of the data to alert decision-makers and call attention to the most important information of the moment; and
  • Propagation: automating the transfer of data from the plant or rig floor up to enterprise-level systems or vice versa.

Emergence of EMI for oil and gas

One leading oil and gas company recently described the issues with its upstream operations. These included:

  • Large-scale operations over a wide geographic area. More than 300 pumping wells connected to 30 processing facilities were consolidated through a single pipeline to one of the company’s largest refineries. This complex was the primary crude source for the refinery.
  • A complex, multivendor process and data management environment. The process was controlled through a combination of data acquisition systems from two vendors. Real-time data were collected from multiple sensors, programmable logic controllers (PLCs), and instruments and given to 15 historians with one vendor, with a “master” historian from another vendor consolidating the data. Additional databases contained laboratory test results, maintenance and failure data, and other process-related information, while a business system managed production data.
  • Multiple user roles with different needs. Direct process management was visualized with different displays for the operators and supervisors. There were more than 100 users of these displays, and the screen designs were difficult to modify. Unfortunately, the systems did not offer adequate analytics, and none were incorporated into the designs.

Process engineers could only apply statistical techniques after a time-consuming data collection process involving querying multiple databases, reading text files, and copy-pasting into Excel files, followed by a multistep manual data aggregation process. There was no real-time application of analytics and no way to routinely analyze data from these sources, much less do so in real time. Unit and corporate managers did not have a combined view of process operations, events, problem-solving, and production data.

Next steps

The company recognized that it was not providing its work force with the tools it needed to better understand and manage its critical processes and also that it was not effectively leveraging problem-solving knowledge accumulated over time. The focus shifted to evaluating systems that provide the functions and services delivered by EMI-capable applications as the best path to an effective solution.

Key issues in achieving EMI

Integration. The first two functions of EMI — aggregation and contextualization — require integration with multiple systems and databases. Upstream information architectures are similar to those found in large-scale processing facilities: data collection from sensors, PLCs, and online instruments feeding historians; function-specific data management systems for offline testing and production data; and data acquisition for operator-level visualization and control. This is an inherently complex environment that presents difficult integration problems. While it is possible to use a single vendor’s application to perform many EMI functions, it requires funneling all the needed data through that application and integrating multiple complex databases that have drastically different internal structures and storage techniques. The work required to aggregate time-series, test-sample, and summarized data demands a high level of expertise and constant monitoring.

A new approach has recently become available that greatly reduces the complexity and problems associated with data integration. The EMI systems gaining traction are vendor- and database-agnostic, providing a stable, uniform layer of analytics and intelligence and delivering role-specific content. This approach also avoids adding another large data repository that duplicates existing data. Instead, its direct queries to the multiple data sources allow aggregation to take place at a higher and simpler level. Finally, industry-standard data integration technologies simplify integration and implementation and make the resulting system more stable and sustainable.

Analytics. The oil and gas industry employs a number of sophisticated analytics techniques for process optimization, resource allocation, and other applications. However, sophisticated analytics (statistical models, optimization techniques, neural networks, etc.) require a high level of training and are most often targeted to very specific areas where they deliver high value. These techniques are not so appropriate for the data available from ongoing operations or the needs of a wide range of users.

A better approach is to choose analytics that have known capabilities for detecting changes in processes with minimal false-positives and that are comparatively easy to interpret and communicate. Techniques such as statistical process control (SPC) offer multiple levels of sensitivity to changes, clear and easy-to-interpret graphical presentation, and a known rate of false-positive to missed signals. The application of SPC techniques along with fixed limits derived from process knowledge requires little specialized training, is very robust (adapting to a wide range of data conditions), and is well understood in the industry.

Visualization and propagation. Differences in responsibilities between operators, engineers, and managers require different information presented in different formats. Operators need “real” real-time data and a clear indication that action is necessary. Engineers focus on solving problems and need access to data with longer time horizons. Managers want a broad view of the overall process and data summarized into key performance indicators. It is critical that all the data used to provision content for each role are from the same sources and used for the same purposes. Another common factor is the need for visual and graphical presentation to rapidly communicate complex information in a form that can be interpreted quickly and accurately.

The geographic range of oil and gas operations requires propagation methodology that can keep a widely distributed work force in touch and looking at the same information. Everyone responsible for the operation needs to see “a single version of the truth,” making it possible to instantly collaborate to solve critical problems.

Expected benefits

Oil and gas companies are starting to recognize the value EMI can deliver:

  • Advanced equipment and process monitoring that improves operation and maintenance of assets, resulting in improved reliability and higher uptime;
  • Real-time dashboards with underlying analytics that dramatically improve operational decision-making by pointing out anomalies early on for frontline operators to respond to in real time; and
  • Evaluation of those anomalies by second-line engineers to rapidly determine underlying economic benefits and prioritize based on financial impact.

And as operators and engineers move up the EMI learning curve, rules-based monitoring becomes a logical next step for many. The addition of knowledge capture can create a true competitive advantage that can be leveraged across a business unit, the corporation, or even shared with business partners.

For leading oil and gas companies, these EMI-based differentiators often can determine winners and losers in today’s increasingly competitive economic landscape.