Every brownfield project must have a starting point. Engineers need to have existing field conditions communicated to them as clearly as possible. Demolition plans, tie-in points, system revamps, temporary structure locations and temporarily installed or removed equipment – all are dependent upon the accurate depiction of existing conditions and how those conditions relate to the project. This allows the best quality of documentation to be provided to the field construction team, and can have a direct impact on project cost and schedule.
In the days of manual drafting, sepia prints or mylars of the original drawings were created and used as this starting point. As 2D CAD use became prevalent, drawings were digitised or scanned into CAD formats for use as the starting point for subsequent design work. With the advent of 3D CAD, a 3D as-built model had to be created.
However, technological advances have put more automation and improvement
into the process of creating an accurate 3D as-built model.
Data capture and display
The use of as-built data capture employs technology in two primary areas:
- The actual recording of the field condition.
- The conversion or presentation of the collected data for use in creating the as-built model.
Field capture technologies primarily include stereo cameras that take photographs (photogrammetry) or laser scanners, which fire a laser beam at an object and record the ‘hits’ on the object in a database.
The use of technology in the conversion and presentation involves taking the raw field data and putting it into a form useful for design work. Stereo photographs are converted to highly precise digital images, which can be used for measurement or to ‘trace’ out a 3D model in a CAD file.
High-fidelity stereo cameras are used to photograph plant objects from multiple angles. Key points, or ‘targets’ are placed on equipment and other items at precisely measured locations and identified in each photo-pair scene. The camera is positioned at a known, recorded location within the plant. Other known information about the camera lens – its settings, imperfections, etc – is also recorded. In many cases, high-intensity lighting is required to adequately illuminate plant objects and the survey tags.
Under ideal conditions, the survey and photography require a smaller team than does manual data collection, and time spent onsite is about 50 per cent that of manual as-built production.
Laser technology has become inexpensive enough to be used in a host of different ways. The use of lasers for generating 3D images of plants or other objects is often called ‘lasergrammetry’ or ‘laser scanning’, and emerging terms for this technique are ‘visual survey’ or ‘high-definition survey’.
Instead of registering photographic data on film, laser beams travel to an object and back to the scanner. The actual 3D location of the ‘hit’’ is registered in a database. Multiple hits can occur within a small distance, giving an actual point contour of the surface of the object. This results in a ‘cloud of points’ that map out a complete surface, and multiple laser scans can be combined to generate a 3D cloud showing all sides and facets of an object.
Using laser scan output
With minor conversion to the digitally collected data, the point cloud can be referenced into engineering design software such as the Intergraph Plant Design System (PDS) and used as a background for design work. Using special ‘plug-in’ applications, the point cloud can be turned on or off, filtered, its resolution adjusted or clipped to focus on an object or an area of interest. BitWyse provides an application called LASERGen, as does Leica Geosystems HDS, formerly Cyra, with the CloudWorx plug-in, that prepare point cloud data for use in MicroStation or AutoCAD.
All of these approaches are used – with basic viewing being the most common, followed by manual modelling and then automated modelling.
One size does not fit all
There are two prevalent technologies used in laser scanning today: LIDAR technology and continuous wave or phase-based technology.
Using LIDAR or ‘time-of-flight’ laser technology, shapes and distances of objects are calculated based on the time of flight of a short laser pulse from the scanner to the object and reception of the received echo. LIDAR systems are able to scan relatively large areas at distances of up to 100 metres with high accuracy.
Continuous wave technology is an alternative to time-of-flight. Rather than pulsing, a continuous laser beam of a controlled waveform is transmitted to the object to be scanned. The amount of data recorded is extensive, and the scan distance is limited to 20–40 metres. This results in an extremely dense point cloud yielding a high level of detail.
Regardless of which technology is used, the multiple datasets are processed and combined in a process called ‘registration’ so that measurements and modelling can take place across multiple scanned areas. The registration process can be performed during the survey, allowing the survey team to ensure quality control in the field.
This is an advantage over older photogrammetry methods because field mistakes are minimised, and the quality of the deliverable is greatly increased.
The ‘as-building’ process
The survey is the critical component of the process, regardless of technology. The field survey team canvasses the site, studying its layout visually and using available drawings. A survey and capture methodology is developed. The plant is then broken down into multiple zones, and each zone is photographed multiple times from many different angles and locations to thoroughly document its characteristics.
There are several ways to integrate as-built data with design tools. Systems can produce fully intelligent PDS or Frameworks Plus models, or semi-intelligent or ‘dumb’ MicroStation and AutoCAD models. The project budget and intended use of the captured data determine which should be used. Most laser scanner vendors provide software that can view, render and produce 3D models of point clouds.
However, most engineering, procurement and construction (EPC) companies prefer to save the time and expense of requiring a model from the scanning vendor, and they simply use the point cloud data as a background ‘obstruction model’ in the plant designing system. After the individual laser scans are combined into a single point cloud, a modeller uses a special viewing tool that loads the digital dataset in PDS.
While the selection of the technology is important, the most important aspect for the success of any as-built capture is the specification of the task itself. As much care and discretion should be used in selection of the as-built data capture contractor as would be used for any other precision project. Plant owners and EPC companies must allow sufficient time to specify the area to be captured, the precision and tolerances of the data capture and the exact deliverables to be provided by the selected contractor.
Qualifications, certifications and experience – these are key to project success and the optimal use of captured data in the design process.
Keith Denton is with Intergraph Process Power & Marine. For more information, visit ppm.intergraph.com