Thirty-seven miles from Salvador on the northeast coast of Brazil, more than 250,000 barrels per day of gasoline, propane, kerosene and other fuels are processed at the Landulpho Alves Refinery (RLAM)--the “Giant of Bahia.” An operational unit of Petrobras that was originally constructed in 1949, the refinery is now undergoing a $1.3 billion expansion.

As part of the project, Petrobras is making major renovations to one of its oil platforms including assembling and installing a new diesel hydrotreatment unit, an electric substation and a new local control room. The company has enlisted São Paulo-based Engevix, a consulting engineering firm, to manage the renovation work.

Working with blueprints that were old and outdated, the engineers began the challenge of planning new lines. The project team also needed to evaluate the route of the new lines and decide whether new pipe racks needed to be built or whether existing racks and foundations could be used. They knew a number of discrepancies would occur when comparing their designs to the actual environment; however, if any new piping didn’t fit perfectly with the existing structures, plant closures and downtime could result. “Every day of downtime to an offshore platform costs an estimated $2.8 million in lost revenue,” says Marcos Migliorelli, director of technologies for Resolution Engineering, a Petrobras subcontractor. “The designers needed a solution that would not dramatically increase engineering costs but would provide the necessary information to ensure an exact fit of the new piping to the old.”

The models were sent to Engevix for analysis the same afternoon the scans were acquired.

Fused Imagery

In a traditional process, engineers would create a computer-generated as-built model of the structure using LiDAR points and photographs. The new design would then be imported into the model to try to find the clashes. However, this process is time consuming and prone to errors. Migliorelli suggested that the team evaluate Life Dimensional 3D (LD3), a new technology developed by Salt Lake City-based InteliSum Inc. that merges LiDAR imaging, digital image data and ground control station (GCS) positioning.

LD3 uses commercially available LiDAR scanners retrofitted with a proprietary 3D camera, the InteliCamera, to simultaneously capture scans and digital imagery. This imagery is fused with--not splattered over--the LiDAR points so that the pixels of the image contain information and the data between the points are retained. Using a patented method of ray tracing, the technology also accurately calculates the data between the points to ensure that every pixel contains accurate dimensions. The accompanying software, LD3 Studio, imports the fused data from the scanner and the camera and exports it to an environment from which the user can “mine” data such as measurements, elevations, distances, grades and bearings. The survey tools enter the data into a field-book-type form, which includes all of the same information that traditional surveying provides.

Engevix determined that this technology would help relieve the concern over gathering all of the necessary details on the existing systems and connecting new pipelines to the old ones while also streamlining the overall project. “The project bid was the same as the cost of manually measuring the 128 interconnections using traditional methods, but the LD3 technology provided the opportunity to complete the project much faster and with a lesser chance for downtime,” Migliorelli explains.

After conducting a successful pilot project with Petrobras on another platform, Engevix decided to obtain scans of the RLAM refinery so that the firm could study different engineered proposals for feasibility. The scan area was nearly 4,000 feet long by 660 feet wide, and all scan data had to be within a 2-millimeter accuracy. Overall, the project required 300 scans. Rather than perform the scans and post-processing work themselves with the new technology, the engineers opted to rely on InteliSum’s expertise.

The scan team, which consisted of between two and four trained and certified engineers, scanned the area with an InteliCamera-enhanced Z+F 5006 scanner. The scan contained all of the surface and visual data and delivered an instant as-built of the environment. Within minutes of scanning, the design team had imported their designs into the LD3 environment. Using a tool called Target Link in the LD3 Imager software, the team was able to easily register the targets in the scene to the as-built survey simply by clicking on the center of each target. In one case, nearly 120 adjustments needed to be made to the design. By the end of the afternoon, the team had a list of design changes that were needed to ensure that the new piping would align with the existing structures.

The models helped the project team ensure that the new piping would align with the existing structures.

CAD-Ready Models

Once the design engineers were able to see their errors quickly, they also wanted to be able to adjust their designs and create accurate CAD-ready piping models inside LD3. To accommodate this request, Stan Coleby, InteliSum’s vice president of engineering, led his software development team to create an extensive new plug-in for the software environment. Working from a proof-of-concept plan made nearly a year earlier by InteliSum employees, Coleby and his team created LD3 Builder.

The Builder tool allows users to model pipes by selecting points in the visual 3D environment. Because users are able to compare their models against the real-world scenario, they can be more accurate with their designs. Also, design engineers only need to model the locations of the desired changes and not the entire scan site. As a result, only a small fraction of the scan data needs to be modeled. The plug-in also contains a Best Fit tool that aligns the models to the scan data to make the model more accurate. The resulting model is CAD-ready and can even be pulled into CAD programs with the real-world textures.

LD3 Builder is now being used extensively in the RLAM design and engineering processes. The designers are using the models to gather accurate dimensions in Bentley MicroStation, and these data are sent to the shop for fabrication. “These tools make the process from design to fabrication much tighter and less prone to fabrication errors,” Migliorelli says.

Another creative use for the technology is that as the fabrication shop creates the individual components, they are scanned and validated against the LD3 environment. Migliorelli explains it this way: “The worst-case scenario is that the new pipe is sent out to the platform only to find out it doesn’t fit. This will cost the company an extra transportation cycle and will keep the platform closed for an extended period of time. By following this work flow, the team can all but eliminate that risk.”

Speed and Quality Control

The RLAM scanning project has been enlarged to encompass other platforms, and the scan work is still underway. Two separate scan teams capture the scans during the day, and these data are sent to InteliSum’s Salt Lake City office each evening for post processing and analysis by InteliSum’s engineers. The post-processed scans are then forwarded to Engevix to be used in the construction process. Through this project, several Brazilian locals have been trained as scan team members. This team, known as InteliSum Brazil, will remain in Brazil working on various platforms.

When the RLAM refinery expansion is finished, the new and improved “Giant of Bahia” will be able to produce a greater output of oil at a more efficient rate. Thanks to the new scanning and modeling technology, the process to complete the expansion has been streamlined, as well.

“Without a doubt, the use of LD3 has given the best results we have ever received,” says Milton Hulpan Pereira, director of operations for Engevix. “It has made it faster and more cost-effective to create detailed designs and solutions--all of this with the added benefit of guaranteed quality control.”

A scan of the Lockheed Martin plant room.

Sidebar: Streamlined Plant Modeling

Although InteliSum and InteliSum Brazil provided the services for the RLAM project, InteliSum is primarily a hardware/software company. Many engineering and surveying firms prefer to handle the scanning work themselves.

One example is CH2M Hill, which recently used LD3 to complete a large pilot project for Lockheed Martin in Moorestown, N.J. The project consisted of scanning a plant room measuring approximately 861 square feet that contains boilers, chillers and other equipment connected with numerous large and small pipe runs. The project goals were to capture an as-built LD3 model of the room and create accurate 3D models as well as 2D planned views of the room. The project team included CH2M Hill’s John Brown and Josh Taylor, who acted as subcontractors for the scan work, and Louie Bush of Entrix, an InteliSum strategic partner, who provided onsite customer service support.

On January 7, 2009, the scan team prepared the room for scanning by conducting a site walk-through and setting up targets in the environment. Using a Z+F 5006 scanner with an InteliCamera, the team completed two scans by that evening.

A model of the Lockheed Martin plant room.

The team returned at 7:45 the next morning and completed the remaining scan work before noon. The scanned data were processed and modeled at InteliSum and then returned to the scan team. By 2:45 p.m. on January 8, all of the data were registered. By 7 o’clock the next morning, the team had collaborated on how to use the models in a presentation meeting to Lockheed Martin, the post processing was completed, and all of the desired 3D modeling in the environment was finished. By 11 a.m. on January 9, the data were included in a presentation to Lockheed Martin.

From preparation to presentation, the entire process took less than 48 hours--including three hours to prepare the site, six hours to scan, two hours to register the data, one hour to plan and collaborate and 11 hours to post process the data and create the 3D models. “The coordination between CH2M Hill and InteliSum on this project allowed our team to deliver a high-quality product to the client relatively quickly and provide them with data that would have taken nearly two months to collect using conventional surveying methods,” says Chris Royak, PLS, survey manager for CH2M Hill. “We have actually used the technology on several projects and plan to continue to do so. When used in the right application, [it adds] value to the project.”