Industry voices speak to oil and gas projects future

Multi-discipline efforts converge in 3-D reality models, even for existing facilities. Four steps for successfully using a data platform are highlighted.

By Bob Vavra, CFE Media February 10, 2017

The hum of a drone can just barely be heard floating above a working oil refinery. A camera is attached to the drone, generating the digital photography needed to build a 3-D reality model, a virtual representation of an existing facility.

Once on the ground, the photos are converted into the reality meshes that culminate in the plant’s 3-D model. Other photography and digital inputs will make the model an immersive environment that mirrors the plant’s most important means of production.

With fluctuations in oil and gas prices slowing investment in industry capital projects, owner-operators are emphasizing increased efficiency of their existing assets. Making better use of digitization is one way of addressing the inherent challenges. As upstream, midstream, and downstream oil and gas industry companies make more use of 3-D reality modeling, what are they getting into and what is their mind-set?

"Business as usual isn’t good enough," said Alec Miljkovic, a Shell project manager working on its capital project-delivery model, ProjectVantage. "We have to stretch ourselves."

Shell’s ProjectVantage was just one of several oil-and-gas industry initiatives featured at Bentley Systems’ Year in Infrastructure 2016 Conference, held in London in November of last year. Several speakers at the event’s oil, gas, and chemicals forum acknowledged a gap in capabilities for brownfield and greenfield projects, as well as a need for increased efficiency in turning digital assets into operational models that can succeed in daily use.

Shell’s ProjectVantage simplifies work packaging for engineering, construction and installation while adhering to the Construction Industry Institute’s advanced work packaging methodology.

Shell says it selected ProjectWise ConstructSim as the foundation of its construction management solution to support its ability to visually plan and execute work safely, avoid premature mobilization and downtime and remove constraints that can impact safety, logistics, materials, labor availability, permits and documentation. Moreover, the solution supports engineering readiness through advanced work packaging to improve constructability and improve field productivity. 

Criteria for success

"We talk about outcome-based engineering. Efficient execution also is an outcome," Miljkovic said. In other words, the criteria for success must include more than whether an asset is in production."

Miljkovic added, "A project is something you tolerate to get increased capacity, but the project has to be executed efficiently. We must focus on where the value leakage is, and how we address those leaks. With data-centric solutions, we get improved collaboration between contractors and suppliers."

Success calls for collaboration among enterprise functions, production units, and even regions, said Ken Douglas, BP information technology and services (IT&S) director of global projects. "It wasn’t that long ago that we ran our capital projects on a business-unit-oriented basis. You could visit the same capital projects in two parts of the world, and they would be very different."

In 2009, BP established its Global Projects Organization to bring standardization to the company’s design-and-build projects. When the group started its work, cloud-based computing was just beginning to have an impact. Its proliferation and the advent of the Industrial Internet of Things (IIoT), particularly for better asset management, changed the project standardization equation once again.

In the future, "data considerations will drive decisions we make about standardization," Douglas said. 

Beauty and the industrial beast

Videos of 3-D mesh designs bring out industrial infrastructure’s beauty in a much different way than does, for example, black-and-white photography. Moreover, users "fly" through models to observe equipment and peer into processes. It has the facility of a video game—except that’s not what’s important.

"Even though reality modeling is beautiful, it’s not really the key point. This is a vehicle for communicating information," said Phil Christensen, Bentley Systems vice president for analytical modeling. "The interesting thing is the crossover to engineering. Yes, you can get operational data, but to help people make decisions, the real job is analysis and simulation."

David Philp, consultancy manager for AECOM’s oil and gas business, said that to take full advantage of digital models, the discussion must extend beyond design and construction. AECOM is a provider of engineering consulting and project management services for infrastructure projects.

Within operations, Philp said, "We’re an output-based industry, and we need to focus more on outcomes. We’re thinking about how to improve asset performance, and we’re learning from real-time data. We can start to evaluate how the plant is going to perform right from the very beginning."

It’s important to note, Philp added, that operational expenditures (OpEx) are about 80% of a project’s total lifecycle cost, even though capital expenditure (CapEx) projections tend to determine which projects are approved. "We’ve got to think more about OpEx right at the very beginning of projects," Philp said. "We’ve got to think about organizational performance. We’ve got to start by thinking about asset management and maintenance strategy."

Outcome-specific data

Management’s desire to make greater use of complex models in day-to-day operations reflects the value of data associated with valuable oil and gas commodities. "Asset management is not about managing assets; it’s about creating value," said Ian Bush of Black & Veatch, a global engineering, procurement, and construction (EPC) contractor active in energy markets. "You tend to value those assets in economic terms; it’s really about outcome-based engineering. It’s about having the right information throughout the life of that asset."

In the future, asset values will be held on balance sheets in ways much different than they are today, Bush said. "If you’re not maintaining information about your assets, you’re stripping your company. You’re actually causing a devaluation in your company."

Interest in the right use of modeling for installations in production has led to the acronym, BIM, for business-information modeling. "BIM is all about creating and maintaining a virtual asset," Bush said. "We’re delivering two outcomes. One is a physical asset; the other is a digital asset. We ought to be handing over digital information that is meaningful."

Digital and physical asset convergence is driven by the need to extend assets’ useful lives through better resource maintenance and management. Modeling for existing operations has evolved slowly, but mesh technology’s growth and development means that drones now efficiently create as-operating models of a brownfield facility. Cloud computing means the computing resources needed are easier to come by.

"Our infrastructure is living well beyond its initial projected life. The only way we can close the deficit between what we need and what we can afford is by getting smarter. We need intelligent infrastructure," said Bhupinder Singh, Bentley Systems chief product officer. 

How smart is it?

Intelligence can be embodied in an as-built model generated with today’s 3-D technologies. "Data is vastly more valuable when it’s live and connected," Singh noted. "We’re talking about taking a digitally created reality mesh and bringing it into the construction phase, and then the design file can be taken into actual construction. Then we can compare what was designed with the as-built model."

One aspect of data’s power is that it enables collaboration, said Christensen. "It’s not just about crunching the numbers, it’s about sharing that number-crunching with your colleagues."

Christensen cited an oil and gas company with 265 platforms in the Indian Ocean, all of which had reached the limit of their 25-year design life. Data modeling of the platforms captured each one’s actual structural integrity as the basis for determining each platform’s future.

"Once the platform was examined," Christensen said, "they could decide what to do next based on the structural analysis rather than just on an assumed parameter."

The next step is sharing data across the enterprise, and that’s where the shoe still sometimes pinches. "Some data is on the desktop; some things are in the cloud. But there still is way too much manual data," said Christensen. "It is really about extending asset life and improving performance. You’re making decisions: Do you decommission the asset, try to limp by, or do you recommission?"

The means exist to cost-effectively model the facility, to use the cloud advantageously, and to incorporate an account of all the assets and resources involved into the base model and thereby embody as-built and operational data in a single version of the truth. But there is one more crucial step in the process—the operators.

Christensen cut to the chase on this issue. "You measure the success of software by whether people really use it," he said. "If people find it too hard, they’ll just send an email." 

Four steps to success

There are four steps, Eckard Eberle, CEO, Siemens Process Control, said in a conference keynote, for successful use of a common data platform. The four steps are:

1. Integrated engineering: Eberle cited Tulip Oil in Germany as a company that had integrated its processes from the acquisition of feed stock to production. "They use digitized maintenance, repairs, and operations," he said. "They lowered CapEx and OpEx costs by about 20% and shortened commissioning time by about 70%."

2. Commissioning: "This is most expensive time we have," Eberle said. "All the equipment is paid for, but nothing is being produced. Your personnel have to be prepared for commissioning." At Total Oil’s platform in Angola, immersive operator training, including using an avatar, brought situational training to the workers. "They will be able to run processes they have trained for in simulation," Eberle said.

3. Optimization: "We have enough information. The question is, how do you collect this data and put in an environment and a dashboard so someone really can work with it," said Eberle. By optimizing data, Eberle said one operator virtually eliminated overdue safety orders at one of its operations. "No matter what the source of the data, there was a significant gain in plant reliability."

4. Plant lifecycle management: The virtual reality (VR) screens are a practical way to combine the 3-D modeling and operational data to better guide repair and maintenance operations. "We are enriching the virtual model with real data," Eberle said. "Bringing this virtual information into the plant also is available on site when you go into the plant. You can provide wearable VR work instructions, and then you can guide people to what they have to do." 

Reality modeling goes mainstream

Reality modeling is become a mainstream technology for infrastructure project and industrial asset performance, said Bentley Systems’ CEO Greg Bentley at the company’s Year in Infrastructure 2016 Conference in London this past November.

Reality modeling has advanced to the point where engineers, geospatial professionals and owner-operators can contemplate having the "as-operated" conditions of infrastructure assets continuously captured for creating reality meshes and producing attention-grabbing visuals.

Laser scanners are widely used to capture an object’s digital profile because they are fast, versatile and accurate. However, the large point cloud datasets generated from laser scanning needs to be processed to convert them into manageable 3-D reality models.

Photogrammetry, an alternative approach to building a reality model, is gaining attention by leveraging digital photography, ubiquitous location information and endless computing power in the cloud. Photogrammetry is software that automatically turns digital photographs into a compact 3-D reality model.

In 2015, Bentley saw this trend emerging and subsequently acquired ContextCapture software, which generates a 3-D reality mesh that turns the photography into a reality model.

Greg Bentley says photogrammetry software supplements more expensive laser scanning, and in many instances, will replace it. For example, photography can add more detail to an existing model or be used to document change. At the conference, Bentley Systems announced that that the latest version of its ContextCapture software combines laser-scanning-generated point-clouds with photos, as "hybrid inputs," for reconstruction into a compact reality mesh format.

"Embedded" ContextCapture software licensing is rapidly expanding from the specialized processing centers of leading 3-D city mapping specialists to leading industrial drone suppliers, it was further noted.

The conference technology keynote highlighted performance gains in immersive viewing of reality meshes, including access to geo-coordinated digital engineering model information, from any browser-ready device.

"Drones, mixed-reality devices, the Industrial Internet of Things (IIoT)—and in fact, "digital natives"—will converge to advance infrastructure engineers, project delivery and asset performance," Bentley said.

Founded in 1984, Bentley Systems has more than 3,000 colleagues in over 50 countries, more than $600 million in annual revenues, and since 2009 has invested more than $1 billion in research, development and acquisitions.

Final words

Companies embrace modeling technologies for their operational value. Compelling reasons exist for those who have to-date been skeptical to make the leap soon. "It’s doable today. It’s not science fiction; it’s fact," said Singh. "When you have business processes that are paper-centric, you’re losing something. If you send an inspector into the field with a paper form, you’re losing something."

Bob Vavra, content manager, Plant Engineering, CFE Media,

Original content can be found at Oil and Gas Engineering.