The ultimate, perfected vision of intelligent operations is a closed loop with the reservoir simulation as the broad base of the system, ongoing operating parameters all captured in real time and fed into the simulation, and the continuously updated simulation becoming the driver for optimization decisions.

Those closest to the task of bringing a closed-loop system into existence view the needed reservoir-level technology as attainable and not far off. While that technology is essential, however, they see the biggest challenges as the adaptation of people and the development of work processes needed to make the vision a reality. The key is to foster a thoroughly collaborative work environment involving all stakeholders and focused on the information generated by the loop. Visualization technology plays an important role in the presentation of that information.

"The loop of real-time drilling - planning a well, starting to drill it, getting data in real time, and using it to update your model and position the well in the right part of the reservoir - is quite well established in the industry and working very well," said Satish Pai, vice president of technologies for Schlumberger Oilfield Services. "Real-time fracturing is also pretty well established. And production monitoring, where you get data from producing wells and identify those that are getting into trouble or going down and put in a remedial plan, is likewise a loop becoming quite well established. So many of these smaller sub-loops are actually being put into place now.

"Very often, it's not the technology at issue but putting the people and processes in place. As these loops start to be completed and established, we then move to the sort of Holy Grail loop, which is to simulate and manage your reservoir proactively in real time. That's where I think the industry is going."

Pai noted the big progress in reservoir modeling and simulation software, which has contributed to the effective use of data from the various kinds of intelligent hardware now placed in wells.

"This was not the case 5 to 10 years ago with our complex seismic interpretation software," he said. "It is a bit like Yahoo or Google. When you build software today it has to be enabled for real-time data feeds. Petrel reservoir modeling software is set up to accept data in real time, which allows you, while drilling, to look at where your well is in the reservoir, adjust the model and steer the well accordingly. We've seen the situation where a reservoir model has been updated three times in one day, with the asset team each time comparing the drilling progress and expected results to the predrill estimates. The result was a very productive well."

Doug Meikle, Halliburton Energy Services Group vice president for the Landmark and Project Management product service line, said, "Intelligent operations (IO) solutions need to overcome not only technological but also organizational and operational constraints to achieving sustained optimization. Closing the big loop so that IO enables an optimizing system will involve engaging people to collaborate in the production workflows as and when they are required to provide their input. Technology and data integration for IO can make this possible, but re-engineering of operational workflows and helping stakeholders adapt to the change involved is equally essential. To do this, operators need to be able to communicate a strong vision of what is involved, clearly define the benefits of doing so, and provide a pragmatic roadmap of how and when they will get there." The more direct closed-loop optimization, Meikle said, involves implementing system-wide Advanced Process Automation (APA).

"Although well-proven downstream, where the feedstock is relatively stable, the dynamic nature of upstream operations and the degree of uncertainty involved has prevented other than partial adoption of APA," he said. "This will be overcome when IO can incorporate technology capable of accurate prediction of performance and behavioral characteristics from structured and unstructured production data. This capability will provide operators sufficient advance notification of changing conditions, thus creating a window that will permit them to intervene. This is the ultimate meaning of management-by-exception."

The implementation of real-time drilling centers has been successful in enabling projects that are relatively short, highly intense and of high value, as well as in leveraging learning across multiple projects, Meikle explained.

Eric Deliac, president of eastern hemisphere operations for Petris Technology, said, "The value of simple tools that enhance collaboration between individuals from separate organizations deserves strong emphasis. This is likewise true of those tools that facilitate the visualization of performance/trend criteria - as well as importance and hierarchies - in large data sets. Used together, these tools can make complex datasets highly visible and simple to analyze. The resulting actions and plans are logically derived from the visualization tool, with the main issues prioritized so that collaborative work is well focused on eliminating the most significant problems. This type of approach creates a shared culture of teamwork and high performance, which in turn results in intelligence.

"The old days of the operations belonging to one single operating oil company are over, and it is increasingly imperative to gather contributions from partners, local government institutions, contractors, and sometimes institutes or universities. This raises some very interesting challenges about the generation of intelligence. More and more technology development is required from contractors/suppliers, but the profit or return is still quite unequally shared between operators and contractors. Is this sustainable?

"Finally, one cannot emphasize enough the key role played by the Web and by Internet technologies. Virtual teams have become the de facto standard, with project teams split between various organizations and locations. Remote access to information, applications and knowledge has become fundamental to intelligent operations."

Collaborative workroom environments that have recently emerged, equipped with remotely enabled real-time visualization capabilities, have brought a vital new dimension to the work process.

"We started a number of years ago with reality centers, which are big presentation centers that bring people together around their common data," says industry consultant Bill Bartling. "Those were a big success, but they were still just that, presentation centers, not workrooms, because the projector technologies did not have bulbs bright enough for lights to be on as material was viewed. Today, these technologies are so bright that images can be on the wall in a room that is completely lit. People can work there. So these technologies are really changing the dynamics of how these visualizations are used.

"On a visit to a collaborative work center at Statoil recently, I saw a live picture of a moonpool on the wall as drilling operations proceeded. The team was having live videoconferences with people out on the platforms. So all of a sudden, technology has transformed these rooms from places where you might go once a month to places you go every day. But the cool thing was that they were bringing in the contractors as well - the driller, the mud contractor or whoever was involved with that part of the decision process. So rather than sitting around on an oil rig in your hard-toed shoes covered with drilling mud, you're in a nice comfortable office with your golf shirt on and pointing at a screen. When I started in production geology years ago (with Chevron), I didn't know the name of the company engineer responsible for drilling my wells, let alone any contractor involved. I never met these people."

Mike Hauser, i-field program manager for Chevron, said the critical pillars for implementing closed-loop optimization across the production sphere are artificial intelligence, change management, transformation of the asset decision-making environment and development of an integrated system for asset management.

"In artificial intelligence and understanding, we need to strike the right balance between automated workflows with a closed loop and domains in which human intervention is needed," Hauser said. "We likewise must focus on change management, encompassing communication, stakeholder analysis, the work process, skills assessment, training and transformations in organizations and their roles.

"Furthermore, we should expand the use of collaborative decision support centers today and move toward transformed environments for asset decision-making, both physical and virtual. There is finally a strong need for an integrated asset management system, where numerical models today and graphic and visual models tomorrow can be used to reduce risk and uncertainty by simulating decisions and outcomes before decisions are implemented."

Chevron has joined with the University of Southern California to create a Center for Interactive Smart Oilfield Technologies (CiSoft), which will focus R&D activities on such opportunities, he added.

The challenges of closing the optimization loop are strongly evident in a project Shell has undertaken involving its Sarawak Gas Production System in Malaysia.

"This is a very, very complex system with multiple sources, compositions and production levels that feeds into three liquefied natural gas (LNG) plants," said Pieter Kapteijn, smart fields program manager for Shell Exploration Production. "It almost requires a real-time perspective just to understand what is going on in the system and all the deviations in it to be able to optimize its performance.

"First of all, we are trying to build a very complex model, update it with the right real-time data - not only about production rates and pressures, but also composition - and then take all the steps needed to optimize the system yield into the LNG plants. If you're good at that, you wind up with a number of extra spot cargoes per year. If you're not, you don't have the supply security you wanted. We've discovered that we really need very well-developed optimizers because you can get to a local optimum, but that's not the same as a global, total-system optimum."

Shell is working with a number of educational institutions, including MIT and universities in Australia and The Netherlands, to develop state-of-the-art optimization algorithms for such highly complex problems.

"One of the issues you encounter is that the more you dive in, the more opportunities you see to optimize things at a still larger system level," Kapteijn said. "Now, sometimes your contractual agreements and health, safety and environment boundary conditions don't allow you to do a total system optimization. Different parts of a system may need to be operated under different constraints. But even then, having an overall model to look at the impact of those decisions at a lower system level is always worthwhile to have."

It is safe to say that this tension between what the industry is capable of achieving and what it actually implements in a given case will be a fundamental part of the closed-loop optimization business picture for years to come. No doubt, two of the most frequently asked questions will be, "How far do we go?" and "How far do we go today?"