Bachelour Thesis: As built report on oil platform jacket

Discuss all Oil & Gas Industry related issues here.
Post Reply
YngveStaberg
I have made <0 posts
I have made <0 posts
Posts: 6
Joined: Wed Feb 01, 2012 8:33 am
Full Name: Yngve Staberg
Company Details: Hogskolen i Gjovik
Company Position Title: Student
Country: Norway
Linkedin Profile: No

Bachelour Thesis: As built report on oil platform jacket

Post by YngveStaberg » Thu Feb 02, 2012 9:49 am

Hi!

I'm a 27 year old student at the University College of Gjovik in Norway, currently working on my bachelour thesis with Vitec AS. The jackets are produced by Kvaerner Verdal AS. My goal is to compare an as built scan of an oil platform jacket with measurements done with a TPS. The scanner that I am using is a Leica Scanstation 2, the software i have available at the moment is 3D Reshaper and Cyclone Model, Register.

The measurements I want to compare is the distances from center, from leg to leg. This requires some modelling of the pipes, and measuring from center to center. For a visual on this look on my next post further down.

I have very little experience in using Scanners, and I'm hoping by posting my progress here you guys can give some hints and helps during the process.

Yours Sincerely

Yngve Staberg

YngveStaberg
I have made <0 posts
I have made <0 posts
Posts: 6
Joined: Wed Feb 01, 2012 8:33 am
Full Name: Yngve Staberg
Company Details: Hogskolen i Gjovik
Company Position Title: Student
Country: Norway
Linkedin Profile: No

Re: Bachelour Thesis: As built report on oil platform jacket

Post by YngveStaberg » Thu Feb 02, 2012 10:02 am

I have currently done a test- scan on a smaller jacket. This scan went pretty well, only exception was that the communication between the scanner and PC was interrupted if the scan was to big.

The picture below shows the test scan of the smaller jacket. As you can see,there are 4 legs sticking up from the jacket,where the platform itself will be mounted. This scan consists of two scans, with a cloud density at 2mm pr 10meters. The distance between each leg(pipe) is about 20meters, and the scanner is placed about 50 meters from the object.

Image

YngveStaberg
I have made <0 posts
I have made <0 posts
Posts: 6
Joined: Wed Feb 01, 2012 8:33 am
Full Name: Yngve Staberg
Company Details: Hogskolen i Gjovik
Company Position Title: Student
Country: Norway
Linkedin Profile: No

Re: Bachelour Thesis: As built report on oil platform jacket

Post by YngveStaberg » Thu Feb 02, 2012 10:29 am

In the next stage, I tried some modelling with 3D - Reshaper.I exported the model in to a *.pts file and imported this into 3D Reshaper. Here I used the function "extract shape" and cylinder. Put the diameter in fix mode, atits theoretical diameter of 1.2 meters. The model seems to fit pretty decent at this stage. With the "extraction tolerance " I tried to go as low as I could before the model jumps out of its path. It seems like it does this if it doesnt have enough points to use as reference.

Image

When measuring the distances between alle the legs(pipes), I used the command "measure with mouse", and hatched the boxes for "point on selection" and "vertex / end "

The differences in measurements from the scanner and tps as built in this case was up to 6 mm, which I think i too much. Or is it exptected? Both measurements was done at the same time in the same conditions.

User avatar
Romuald
V.I.P Member
V.I.P Member
Posts: 106
Joined: Wed Mar 12, 2008 10:26 am
Full Name: CLAVE
Company Details: www.urbica.fr
Company Position Title: Project manager
Contact:

Re: Bachelour Thesis: As built report on oil platform jacket

Post by Romuald » Thu Feb 02, 2012 12:18 pm

YngveStaberg wrote: The differences in measurements from the scanner and tps as built in this case was up to 6 mm, which I think i too much. Or is it exptected? Both measurements was done at the same time in the same conditions.
If I understund well, you compare the As built (the clouds of point) to a theoritical 3D model of the rig ?
I can not validate your operation In R-Sheper, I did not use it.
But if you are doing well, 6mm deviation seems for me acceptable. I Have work on a lot of case where deviation was really bigger. Legs rig are submit to so many constraint !
Anyway, be careful when you try go under 5 or 3 mm tolerance. A lot of parameteres should be integrate to validate such precision.

Regards

Romuald

YngveStaberg
I have made <0 posts
I have made <0 posts
Posts: 6
Joined: Wed Feb 01, 2012 8:33 am
Full Name: Yngve Staberg
Company Details: Hogskolen i Gjovik
Company Position Title: Student
Country: Norway
Linkedin Profile: No

Re: Bachelour Thesis: As built report on oil platform jacket

Post by YngveStaberg » Thu Feb 02, 2012 12:22 pm

I'm sorry if I was unclear. But the comparison is between the scanned as-built, and the as - built measured with a totalstation. The TPS measurements are considered as very accurate, with a standard deviation of +-2mm. In my next phase I will try to extract the measurement for Cyclone model, it will be interresting to see how the measurements from the two different softwares compare.

User avatar
Romuald
V.I.P Member
V.I.P Member
Posts: 106
Joined: Wed Mar 12, 2008 10:26 am
Full Name: CLAVE
Company Details: www.urbica.fr
Company Position Title: Project manager
Contact:

Re: Bachelour Thesis: As built report on oil platform jacket

Post by Romuald » Thu Feb 02, 2012 12:46 pm

It is effectively different ;)

Compare unique point survey with total station to 3D clouds is for me a no sense.
You fit Cylinder on thousand points, it is the best to garantee high precision on diameter and direction, but ponctually, you could get deformation, pb of reflectivity... Adding to the factory of precision for each points and difference of precision between the both system... One other point having consequences for your comparaison is the positioning of TPS and 3D leser scanner in the same coordinate system.
Can you developp those points ?

YngveStaberg
I have made <0 posts
I have made <0 posts
Posts: 6
Joined: Wed Feb 01, 2012 8:33 am
Full Name: Yngve Staberg
Company Details: Hogskolen i Gjovik
Company Position Title: Student
Country: Norway
Linkedin Profile: No

Re: Bachelour Thesis: As built report on oil platform jacket

Post by YngveStaberg » Thu Feb 02, 2012 12:54 pm

hmm, that is a strong argument. The circle made from the TPS is based on only 5 measurements, while the scanned one has severly more. Even so I think the difference should not be this big. The TPS measurements are done in one stand, and the scan in two. The constraints seems pretty good. Both are in a local system scale 1 to 1. And the measurements are relative.

Image

Post Reply

Return to “Oil & Gas Industry”