Spatial information – an object’s location, size, shape and relationship to other structures and objects – is an essential component in planning, constructing and operating energy facilities.
In order to support efficient production and transport of oil and gas, energy producers must have accurate, timely information about structures, pipelines and other operating assets. Various forms of spatial data, ranging from 2-D maps and schematics to 3-D drawings, have long been familiar tools for engineers and plant operators.
But gathering spatial information with sufficient accuracy and detail can be costly and time-consuming.
Headquartered in Denver, CH2M has adopted a new approach, using a blend of spatial sensors and software to support work in gas pipelines and facilities. The use of combined technologies produces spatial information and detailed documentation of construction sites, potential pipeline alignments and existing plants.
The company uses this dense flow of information for engineering design, analysis, procurement, construction and project management. Going beyond the traditional spatial information of position and description, the information helps develop answers to the deeper questions of the status and condition of facilities and equipment, how to maintain and repair them, and how to do so safely.
The collected processes that gather, analyze, model and use spatial information for plants and fixed assets are known as “industrial geomatics” (IG). IG serves as the platform to gather and manage information on complex sites and facilities.
For many plant and facility applications at CH2M, traditional surveying and measurement is morphing into a mechanical/industrial 3-D geodatabase. The company is finding immense value in 3-D spatial information – IG data is proving to be at least as valuable in the operations and maintenance aspect as it is for engineering design.
Scanning and surveying a facility for both engineering and operations delivers a twofold yield. By using industrial geomatics, CH2M can quantify and eliminate geospatial unknowns to tell people what they have and where it is located.
Industrial geomatics enable various disciplines to see things from the clients’ perspective. For many clients – IG data goes both to in-house and external recipients – the method for collecting information is unimportant.
For example, by the time a project manager sees the results of a survey, the large volume of field data has been boiled down to information needed for design and analysis.
It’s common for multiple specialties, such as environmental, revegetation or planning, to work off a single data set, extracting similar information for different purposes. By leveraging the ability to have a multi-disciplinary project team work from one database, CH2M HILL can introduce efficiency and cost savings across the board.
For years, CH2M has used 3-D laser scanning as a source for geospatial data, and it remains the tool of choice for many applications. The combination of point cloud data and sophisticated 3-D modeling software provides good results in industrial settings.
Recently, however, a new approach emerged to collect and manage data on large, complex projects. Early in 2014 the company began working with a device known as an imaging rover, which combines surveying and terrestrial close-range photogrammetry. Mounted on a survey rod, the Trimble® V10 imaging rover contains an array of 12 high-resolution, calibrated cameras that captured a 360-degree panoramic image in a single operation.
When integrated with a global navigation satellite system (GNSS) receiver or prism target for a total station, the imaging rover automatically produces 60-megapixel geo-referenced panoramas. The system is controlled using a handheld controller or tablet and captures images and position in the same time (typically less than five seconds) needed to complete a survey measurement using optical or GNSS sensors.
Using dedicated field software, the operator can confirm the image and positions have been correctly captured and recorded before moving to the next point. By collecting multiple panoramic images on a site, crews can capture information to provide complete visual and empirical documentation of the site’s features and conditions.
GNSS or optical-positioning sensors enable images to be quickly related to objects and features that are tied to a geodetic or local coordinate framework. If points with known coordinates are visible, they can be used to tie the panoramas to the known coordinate system.
The imaging rover produces panoramic images by combining images from 12 cameras. The resulting data can be used in much the same way as imagery from airborne or terrestrial cameras. In order to combine multiple panoramas to produce accurate 3-D data and orthomosaics, field crews must capture overlapping images that contain a sufficient number of common, readily identifiable features (known as tie points) from one image to the next.
This is accomplished by controlling the distance between photo stations, i.e., the location of the imaging rover when each panorama is captured.
Field procedures also control the accuracy and resolution of 3-D point clouds produced from the panoramic images. The resolution, which is the distance between individual 3-D points in the point cloud, is a function of camera resolution and distance from the camera to the object. Similarly, the accuracy of a point’s coordinates derived from imagery is related to the distance from the photo stations used to compute it.
Because of the inter-relationship between overlap, accuracy and resolution, a series of simple calculations defines suitable field procedures when collecting images to be used in constructing a point cloud. Prior to entering the site, survey teams can determine the distance from the object and spacing between photo stations that will produce imagery to meet project specifications and produce required deliverables.
Once on site, the crew develops an approach to capture the needed imagery. Because the location of the photo stations can be measured using GNSS or optical methods, it’s not necessary for the stations to be inter-visible as long as requirements for overlap are satisfied.
Positions from Pictures
Information from the imaging rover and positioning sensors is downloaded to Windows-based office software running on desktop computers. Upon import into the office software, the data from each panoramic photo station becomes a calibrated combination of 12 images resulting in a 360° field of view.
By using common tie points to adjust multiple photo stations that see the same object, technicians can develop orthomosiacs, 3-D point clouds and individual points. The once-tedious task of identifying common points among photos – an essential component of photogrammetric processing – can be performed automatically.
When processing is complete, the software provides close-range photogrammetry tools that enable technicians to determine 3-D coordinates for each pixel in the photos. Users can extract positions and dimensions within the images to survey accuracy.
Visualization tools make it possible to move through the site to see objects and points from any photo station viewpoint. The panoramic data can be combined with imagery from other sources, including aerial photographs (from manned or unmanned aircraft), as well as terrestrial imagery captured by total stations.
The imaging rover can be used as a standalone solution or to enhance point cloud or traditional survey data. When working on facility-scanning projects such as refineries and petrochemical facilities, crews can use the imaging rover to supplement scanning data. The unit is placed in a confined or obstructed area to capture details not visible from the scanner, tying the images into the point cloud without the need for additional scanning targets.
In addition to saving the time needed for an additional scan setup, the approach provides images that specialized software can drape over point clouds to produce photo-realistic 3-D models.
Pipeline planning and construction require geospatial information over large areas. Crews use the imaging rover to support the three phases of pipeline construction: preliminary surveys and planning, construction staking, and as-builts.
The surveyors use the imaging rover during the planning phases to supplement aerial data and conventional topographic and location surveys. In some areas, only sparse data is required, as in the case of general topography.
In other areas, it may require sufficient density to design a road crossing or avoid existing features. In areas where dense data is needed, the imaging rover collects a high level of detail in seconds. For example, crews capture photos in areas where conditions might require the pipeline to avoid utilities or other structures.
The panoramic images allow designers to see the site from multiple viewpoints and develop alternative plans without sending crews back to measure additional points.
On long-line pipelines, crews often mobilize hundreds of miles down line. To avoid the cost of multiple trips, CH2M uses imagery to develop points and information in the office. If something is left behind, or if right-of-way teams need data for an adjacent property, it’s likely that the additional information exists in the panoramic photos.
The cost to extract additional information in the office is tiny when compared to the time and expense of returning to the site – even assuming site conditions haven’t changed.
During the construction phase, surveyors use GNSS and total stations for stakeout, generally providing coordinate lists or cut sheets to document their work. The imaging rover provides visual records of points set and overall site conditions. In the event that the surveyors’ work is questioned, the photos can show the as-staked points and confirm their positions.
As pipeline construction proceeds, crews gather as-built data including top-of-pipe measurements to locate the pipeline and its features. Prior to backfilling, a team captures the location of each weld together with quality-control information (photographs, X-rays, serial numbers and more). They use GNSS and the imaging rover to capture locations and panoramic images and then detach the tablet from the survey pole and use it to take detailed snapshots of individual welds and features.
The positions, tablet photographs and panoramic images from the imaging rover are processed into a spatial database for the pipeline, which provides a complete record of the excavation, welds and pre-backfill conditions.
Geo-referenced panoramas save time in two ways. First, imaging significantly reduces the field time needed to capture 3-D information on points and features and can eliminate the need for most return visits. While it’s difficult to create a direct comparison between imaging and conventional survey data collection, it’s safe to say the rovers have reduced field time by 15-20%. The reduced costs in the field easily outweigh additional office time required for image processing.
The second savings comes from the ability of images to freeze an instant in time. Beyond providing pre-backfill pipeline information, photos show the surrounding conditions on a site or plant facility. In addition to enabling additional measurements, the visual data allows teams to “travel back in time” to resolve questions or disputes and support quality control and environmental aspects.
Accurate spatial information enhances safety efforts. Imaging can remove the need for personnel to enter areas that are hazardous or difficult to access. An accurate spatial database, with easy access and readability, supports good planning on operations and maintenance activities that in turn help prevent accidents or rework.
Operators can use IG information to design and execute plant repairs. Retrofits can be planned to exacting detail and verify new machinery fits into a designated space, and design connections to existing equipment. Spatial data helps planners determine the best way to move large items in a facility without moving or damaging equipment already in place.
Geospatial information provides the backbone for CH2M’s industrial geomatics approach to complex projects in oil, gas and other industrial settings. The technology provides information that is timely, correct and delivered in a way that enables large project teams to work off a single consistent data set.
Author: Jason Jung is manager of 3-D laser scanning for CH2M. Jung provides industrial geomatics services for clients in the energy and petrochemical sectors. Licensed as a Professional Land Surveyor, Jung’s experience includes cadastral, control and construction surveying, as well as 15 year’s work in industrial and facilities surveys.