Recently, one large oil company estimated that reliance on spreadsheets and lack of integration among engineering tools consumed as much as 70% of its production engineers' time on routine data manipulation tasks, leaving only 30% for scientific analysis and diagnosis of well performance. To alleviate this unacceptable situation, the company brought in a small team of consultants to identify people, process and technology issues.

They decided to replace spreadsheets with a dynamic surveillance tool, integrate it with a system analysis tool, and link them both into existing corporate databases. Furthermore, they effectively automated two critical engineering processes - daily performance monitoring and production test validation. Suddenly reservoir management professionals could transform huge amounts of raw data into timely, relevant information. They could manage by exception, quickly pinpointing those few wells out of hundreds that required prompt intervention or further investigation. As a result, they began to reverse the traditional 70-30 time split, better leveraging constrained human resources and, ultimately, boosting production from their under-performing wells.

This large operator, like many others today, faces daunting challenges. With rising global demand, declining production, growing data volumes, dwindling resources, mounting regulatory and environmental pressures, exploration and production organizations must dramatically improve the management of their hydrocarbon assets. However, with increasing field instrumentation and hundreds of wells to manage each day, production engineers are swamped with data. Too often they cannot see the forest for the trees. How can energy companies help relieve some of the burden? By better understanding production operation processes within its organizations, by integrating production data systems and software, and by automating day-to-day engineering workflows. Taking these steps will enable more holistic approaches to production system modeling (Figure 1), the goal of which is to optimize multi-phase flow from the completions through the well, surface network and facilities all the way to the point of delivery.

Production operations assessment

An excellent place to begin optimizing production operations is by conducting a current state assessment of targeted producing assets. This requires thorough analysis and mapping of business and technical processes; data flows and bottlenecks from reservoir to market; identifying existing tools and technologies; determining when and how people impact timing and decisions; estimating both the costs and the projected value of new technologies needed to streamline production engineering workflows; and scoping out thorny change management issues.

A full assessment typically starts with interviews of asset management and team supervisors to understand key business goals and known high-level problems that must be addressed. Next, production operations team members are interviewed to determine what exactly they do on a day-to-day basis, who else in the organization they interact with, what tools they use, what works well now, where they encounter pitfalls and barriers, and generally what keeps them awake at night. Based on these input sessions, workflows are mapped, noting gaps, inefficiencies and current best practices. Information collected during interviews is analyzed to develop detailed findings, conclusions and a list of solid recommendations. These are validated by production operations personnel before being set into motion. Potential short-term and longer-term process improvements are ranked on cost and payoff, and an implementation plan created.

The company found that mundane tasks such as locating, exporting, reformatting and quality-controlling data were consuming valuable engineering time, forcing it into a costly reactive mode. Only after a problem developed would a well receive necessary attention. While engineering technicians could, in theory, take some of the drudgery off the production engineers, in reality they spent most of their time putting out fires. New technologies and automated workflows were needed.

Automation of production workflows

Once production operations processes have been carefully mapped and problem areas identified, it is vital to streamline the flow of data among key production engineering tools and begin to convert data into actionable information.

There are two basic ways to accomplish this: through customized integration of current systems and automation of certain repetitive processes, or by replacing commercial stand-alone systems with more advanced integrated solutions. The ideal approach, of course, would be to sort out the data issues first, then select fit-for-purpose technologies, then automate workflows on top of the new systems.
The whole point of workflow automation is to give production engineers a break from trawling through mounds of daily operational data. An automated monitoring system can do much of the dirty work for them, flagging wells that require attention.

For example, consultants working for the company wrote interfaces to link their new surveillance system (Figure 2) with two legacy production databases and a third-party system analysis tool. To minimize custom coding and ease the transition to a more integrated infrastructure in the future, they decoupled workflows from underlying data stores using reconfigurable interfaces. With the new surveillance workflows, production test data were automatically captured and displayed on calibrated IPR plots, enabling rapid, rigorous validation. In addition, actual well performance rates were automatically compared with modeled values, alerting engineers to any deviations from a predefined range. Engineers' efforts became proactive, their energy focused on issues with real bottom-line impact.

Production engineers have been so impressed by the benefits of automating these workflows, the company is planning to utilize so-called "real-time" (or, more accurately, "sub-daily" or "high-frequency") production data as well. While real-time data from SCADA or data historians are not required for automation, combining high-frequency online data with traditional lower-frequency information can further streamline operations and improve business decisions. On the other hand, capturing real-time data without automation can overwhelm production engineers with information they cannot process in a timely manner. Using real-time data as a diagnostic tool after something goes wrong will never help engineers get out reactive mode.

Integration of engineering technology

As noted, adoption of commercial integrated solutions is another aspect of process optimization. The company, for example, is now considering replacing its legacy database system with a common engineering database designed to integrate a full range of drilling, completion, production engineering and economics applications. Why?

Consider the engineering tools necessary to make an ordinary workover decision. If production test data show that a well is producing, say, half the volume of fluid it had the previous week, a production engineer might run test cases on existing nodal analysis models and discover that sanding is the likely culprit. To address the problem, it may be necessary to bring in a coiled tubing unit to clean out the sand. But how much will that cost? And what's the potential payoff of doing that workover? To make an intelligent decision, the engineer needs access to an effective surveillance application, nodal analysis technology, workover design software and economic analysis tools. Passing data back and forth among disparate, stand-alone applications or spreadsheets inevitably bogs down the entire process, costing valuable time and interrupting the engineer's train of thought. This is, after all, just one of possibly hundreds of wells to manage.

Traditionally, "integration" of engineering data has been achieved manually. Recently, however, commercial technologies have become available that provide, for the first time, a unified engineering data structure for an array of integrated engineering applications, with a consistent desktop user interface (Figure 3). Drilling, well services, production and economics data stored in the common database are shared by each of the applications, eliminating tedious and time-consuming transfer and reformatting of data.

As a result, production engineers can seamlessly perform a range of complex analyses at their desktops with considerably greater ease and efficiency. They can proactively assess well performance issues and provide more timely proposals to asset management with all the technical and economic information they need to make better operational decisions.

The impact of automation and integration

Obvious measurable gains come from reducing operating costs - which in mature assets can be considerable - and increasing production through optimization of the whole production "system," from sand face to point of sale. Typical production increases from a systems modeling approach can range from 3% to 10 %, depending on the type of field. Incorporating high-frequency data in an automated workflow can enable the whole system to operate at closer to its optimum limits for longer periods of time, boosting production several more percent.

While approaches to production engineering automation and integration are still in their infancy, real-world projects indicate that return on investment is always many times greater than up-front costs. In one gas-lifted oil field, automation and optimization substantially decreased gas injection, reducing compression costs while increasing oil production. The new system paid out in mere months.

Isn't it time we gave more production engineers a break?

Comments