Managing the Digital Oilfield of the Future

When you drill an oil well today, you get a gusher of data along with the oil and gas. This is because exploration and other drilling equipment is increasingly wired with instruments and networked to monitoring and control systems. And, like the oil itself, that data also must be refined into a useful form that can help users make decisions.

"These days a single drill pipe may be wired and scanning 30,000 times per second to report on temperature, pressure and other variables," said Doug Johnson, production optimization product manager for Halliburton's Landmark Software & Services division. "And the number one place that data goes to in our industries is a Microsoft Excel spreadsheet, but the main challenge is how to get that data into a representative context that we and our customers can use. Of course, there are many differences between data from asset to asset inside individual companies, their facilities and applications."

Johnson delivered his presentation, "The Digital Oilfield of the Future," as part of the Oil & Gas Industry Forum this week at Rockwell Automation's Automation Fair 2009 in Anaheim, Calif.

To help its many oil and gas users and their facilities handle these data fire hoses and operate more efficiently, Johnson reported that Landmark Halliburton recently combined and rebranded three software solutions under one umbrella that it christened DecisionSpace for Production. "DecisionSpace is an integrated framework for data management, data analysis, data mining, application workflow automation and real-time optimization," said Johnson.

DecisionSpace includes Rockwell Automation's FactoryTalk VantagePoint software, which Landmark calls AssetObserver. The combined software trio also includes workflow management software from Engineous, which Landmark dubbed AssetConnect, as well as Rockwell Automation's Pavilion8 software, which Landmark calls AssetSolver.

"Consequently, DecisionSpace gets data from databases that a user already has, adds outbound software tools and then performs both classic workflows with AssetConnect and predictive functions with AssetSolver," explained Johnson.

"The number one destination for data in the oil and gas industry is still the Microsoft Excel spreadsheet." Halliburton Landmark's Doug Johnson explained how the company uses Rockwell Automation's FactoryTalk VantagePoint and Pavilion8 software to manage oilfield data effectively.As a result, users can drag software icons representing individual applications onto "coat hangers," which form an on-screen flowchart framework. This allows the software to map the icons on the coat hanger, automatically execute the connections between them and run as an integrated process.

For example, Landmark Halliburon recently assisted Chevron's Agbami i-field project, located off the coat of Nigeria. Its combined software is integrating production operations workflow on the project's floating oil rig ship, which is tethered to several oil wells and includes full DCS and telemetry functions.

"Starting in December 2008, we used DecisionSpace to break the Agbami project's overall workflow into several individual workflow tasks. Needed functions are fed into workflows to do reservoir surveillance," says Johnson. "For instance, we're doing virtual metering to help perform well tests that are required for validation."

In the past, Johnson added that most of these data engineering tasks were done manually, which was slow and increased human data-entry errors. "And, doing it manually meant that users still couldn't handle all the data they were getting in, which also limits up-time. The Abgami project was one year behind getting to first oil," explained Johnson. "We were able to get our software installed and all our work done before oil started flowing, and now 95% of the project's wells and facilities are performing well at this time."

Likewise, another user operating 15 oil and gas wells in the Gulf of Mexico is using DecisionSpace to balance its production. It employs the software to create proxy versions of its facilities, develop neural networks and then produce real-time models to optimize performance. "Giving users one integrated model of their application also can help them work through the process of dealing with management of change and people issues that they need to address," said Johnson.