Project at a glance
- Client: A North Sea operator (name withheld), through subsea equipment specialist Ashtead Technology
- Sector: Offshore oil and gas, subsea
- Location: North Sea, subsea; data processing in the UK
- Services: Industrial Measurement, Subsea 3D laser scanning support, data processing and registration, ovality and 3d deformation assessment, 2D CAD to point cloud comparison
The brief
Luminous Group were appointed by Ashtead Technology to assist with subsea 3D laser scanning and data processing on a subsea wellhead belonging to a North Sea operator. The work was carried out underwater, on a live asset, by divers.
The objectives were set out in three parts: “To accurately measure ovality of the internal diameter, deformation within the lock-ring profile, surface corrosion and pitting within the 30 inch low pressure housing down to a vertical displacement of 20 inches.”
Those three questions are all about shape. Ovality is the degree to which a nominally circular bore has gone out of round. The lock-ring profile is the machined seat a connector engages with, so its geometry governs whether equipment will land and lock. All three had to be answered from measurement, not inspection video.
The final report was issued in two revisions during July 2017, running to 48 pages.
The challenge
Subsea measurement removes almost every convenience that onshore dimensional control takes for granted. The instrument cannot be adjusted by hand once deployed. The medium between scanner and surface is water, not air, and the working environment is shared with divers, umbilicals and lifting equipment.
Two specific problems shaped the data processing. The first concerned the registration bars, the physical references used to tie separate scans together. As recorded in the report: “The main issue was that the registration bars which were used were only positioned using a small magnet at the top and a large handle sticking out which had been knocked by the divers or cables whilst in several positions. This meant that in some of the scans the bars had to be chopped out of the data before registration because the slight movement would have introduced errors.”
A reference that moves between scans is worse than no reference at all, because it looks authoritative while quietly corrupting the alignment. Recognising it and removing the affected geometry was essential to the result.
The second problem was water clarity. “In general the water was fairly clear but there was a noticeable difference in the scan data directly after a diver had moved the registration bars and kicked up some sediment creating more surface noise.” Disturbed sediment scatters the laser and raises the noise floor of the data, which matters when the measurement is reported in millimetres.
There is also no cheap second attempt. Every return visit to a subsea wellhead carries dive time, vessel time and weather risk. The capture had to be right first time.
Our approach
Scanning was carried out with Newton Labs subsea laser scanners, diver-deployed and operated by Ashtead. Forty-eight scans were taken of the internal well bore.
Our processing was carried out in Geomagic Design X. The first stage was cleaning: manual fencing to isolate the geometry of interest, filter cleaning to strip returns that were noise rather than surface, and surface smoothing. This is where the disturbed-sediment scans and the compromised registration bars were dealt with, before any alignment was attempted.
Registration then ran in stages. Scans were first aligned sequentially using a transform registration, then refined with a global and fine registration across the whole set. The last scans were joined back to the first to close the loop. Closed-loop registration is one of our standard controls: by tying the sequence back on itself, accumulated drift becomes visible rather than piling up at one end of the dataset. A final quality assurance pass applied additional constraints to the registered result.
3D Data Analysis
With a clean, registered model of the bore, we established a frame of reference. The dataset was aligned to structural north using the notched area at the bottom of the pipe, and a datum of zero was set at the top of the pipe flange. Every subsequent measurement was reported against that datum, so the client could relate any figure back to a physical feature on their own asset.
Ovality was assessed on horizontal slices through the registered cloud. Slices were isolated in AutoCAD and traced to 2D line work, so the measured bore profile could be compared directly against design geometry.
This is the argument for scanning over traditional dimensional control, and subsea is where it is sharpest. Conventional surveys verify a limited number of points. Laser scanning creates a full 3D replica of the item, so a dimension can be checked at any point, or the cloud compared against the design model. It also reduces the need for costly re-surveys, because everything is captured first time.
What we found
Ovality was checked at six horizontal sections below the datum, at 65mm, 110mm, 155mm, 235mm, 390mm and 490mm below the zero level at the flange top, plus the datum level itself.
At each level, eight radial dimensions were measured, labelled A to H around the bore. At datum level zero those eight dimensions measured 402, 407, 404, 398, 400, 400, 396 and 400 millimetres. The spread between the smallest and largest is 11mm.
That spread is the point of the exercise. A bore quoted as a single nominal diameter tells an engineer nothing about whether it has gone out of round, or in which direction. Eight radial dimensions at seven levels describe the actual shape of the bore, including its orientation relative to structural north.
Eight well-bore cross sections were produced, and each level was supported by a 2D CAD to point cloud comparison, showing the measured profile against design geometry.
Deliverables
The issued package comprised ovality section and deformation assessment drawings at each measured level, the eight well-bore cross sections, 2D CAD to point cloud comparison views at every level, the registered point cloud, and a definition of terms appendix.
The outcome
A live subsea wellhead was measured to millimetre resolution, in water, by divers, and reported as a described geometry rather than an opinion. The bore shape, its orientation and its condition were all recovered from a single 48 scan capture.
The registered cloud remains a permanent record. If a further dimension is queried during intervention planning, it can be taken from the existing data rather than from another dive.
Traditional dimensional control surveys only verify a limited number of points. Laser scanning creates a full 3D replica of the item.