where expertise comes together - since 1996

Web's Largest Portal of Nondestructive Testing (NDT)
Open Access Database (Conference Proceedings, Articles, News), Exhibition, Forum, Network

All Forum Boards
Technical Discussions >
MFL - Tank Floors
Career Discussions
Job Offers
Job Seeks
Classified Ads
About NDT.net
Articles & News

Iris Inspection Services, Inc.
We are the world’s largest supplier of air cooler, boiler and exchanger tubing inspection services.

1526 views
03:18 May-07-2002

Mark Badrick

Inspection
Bahrain Petroleum Co,
Bahrain,
Joined Oct 1999
11
MFL - Tank Floors

Up to a few years ago, Tank Floor Inspection at this refinery consisted of Visual, UT gauging, vacuum box testing and the cutting of coupons in selected locations. We have now taken advantage of the many service providers of MFL in the Gulf region. We have, however, still maintained the practice of cutting coupons and on several occasions have found discrepancies between what was reported using MFL and the physical evidence of the coupons.

Firstly, are there any readers of this forum who use MFL and just rely on the UT proove-up to verify MFL findings, and secondly is there anybody that does the same as this refinery and cut coupons to verify the MFL?

We appreciate that volume loss is the key and that narrow through wall pits can actually be missed by MFL, but in this instance the volume loss should have been detected.




 
04:19 May-07-2002
Richard Kazares
Re: MFL - Tank Floors I'm sure we have all heard this before: EVERY NDT method has its advantages and DIS-advantages - MFL and UT being among them.

Volumetric readings can easily be "fooled" by local (or global) variations in materical density, layup (e.g. welds), overlaps, pits (particularly small) and other geometric features that render volumetric readings that are not totally useful.

Also, conventional UT readings (while usually accurate at the spot they are taken) can be likened to finding a single "star" in a universe populated by 1,000,000,000 GALAXIES!!!!! (Well almost, anyway). That is: when attempting to find a small pit in a 2 acre tank floor.

A possible "compromise" to this dilemma is to use a high-data-density automated UT method (there are several possibilities - including ours) which at least LOWERS the odds of missing significant data. By "significant" data - I'm referring to indications that actually might result in structural compromise.

Other "global" solutions include the use of Acoustic Emission as a tool to "rate" a tank floor condition - and determine whether to search in minute detail (e.g. automated UT) or to use spot checks - as a function of the rated "condition" of the tank floor. This strategy can be considered in the "Better the odds" category - which provides a relatively low-cost (non-invasive) method of rating "risk."

The bottom line to all this (sorry about the length) - is that looking for small defects in large areas is ALWAYS a statistical issue - the "best" solutions involve INCREASING the odds of finding "real" (significant) "defects" while providing insight into the condition of the structure.

NO NDT, or other inspection METHODS - including selecting and cutting coupons (obviously NOT NDT) - can claim 100% (absolute)accuracy and assurance of finding EVERYTHING - ALL THE TIME. The best anyone (or any method) can do is BETTER the odds

: Up to a few years ago, Tank Floor Inspection at this refinery consisted of Visual, UT gauging, vacuum box testing and thecutting of coupons in selected locations. We have now taken advantage of the many service providers of MFL in the Gulf region. We have, however, still maintained the practice of cutting coupons and on several occasions have found discrepancies between what was reported using MFL and the physical evidence of the coupons.
.
: Firstly, are there any readers of this forum who use MFL and just rely on the UT proove-up to verify MFL findings, and secondly is there anybody that does the same as this refinery and cut coupons to verify the MFL?
.
: We appreciate that volume loss is the key and that narrow through wall pits can actually be missed by MFL, but in this instance the volume loss should have been detected.
.



 
07:07 May-08-2002
Simon
Re: MFL - Tank Floors Do B-scan on areas identified by MFL. This will give you a cross sectional view of the area in question.


: I'm sure we have all heard this before: EVERY NDT method has its advantages and DIS-advantages - MFL and UT being among them.
.
: Volumetric readings can easily be "fooled" by local (or global) variations in materical density, layup (e.g. welds), overlaps, pits (particularly small) and other geometric features that render volumetric readings that are not totally useful.
.
: Also, conventional UT readings (while usually accurate at the spot they are taken) can be likened to finding a single "star" in a universe populated by 1,000,000,000 GALAXIES!!!!! (Well almost, anyway). That is: when attempting to find a small pit in a 2 acre tank floor.
.
: A possible "compromise" to this dilemma is to use a high-data-density automated UT method (there are several possibilities - including ours) which at least LOWERS the odds of missing significant data. By "significant" data - I'mreferring to indications that actually might result in structural compromise.
.
: Other "global" solutions include the use of Acoustic Emission as a tool to "rate" a tank floor condition - and determine whether to search in minute detail (e.g. automated UT) or to use spot checks - as a function of the rated "condition" of the tank floor. This strategy can be considered in the "Better the odds" category - which provides a relatively low-cost (non-invasive) method of rating "risk."
.
: The bottom line to all this (sorry about the length) - is that looking for small defects in large areas is ALWAYS a statistical issue - the "best" solutions involve INCREASING the odds of finding "real" (significant) "defects" while providing insight into the condition of the structure.
.
: NO NDT, or other inspection METHODS - including selecting and cutting coupons (obviously NOT NDT) - can claim 100% (absolute)accuracy and assurance of finding EVERYTHING - ALL THE TIME. The best anyone (or any method) can do is BETTER the odds
.
.
.
.
: : Up to a few years ago, Tank Floor Inspection at this refinery consisted of Visual, UT gauging, vacuum box testing and the cutting of coupons in selected locations. We have now taken advantage of the many service providers of MFL in the Gulf region. We have, however, still maintained the practice of cutting coupons and on several occasions have found discrepancies between what was reported using MFL and the physical evidence of the coupons.
: .
: : Firstly, are there any readers of this forum who use MFL and just rely on the UT proove-up to verify MFL findings, and secondly is there anybody that does the same as this refinery and cut coupons to verify the MFL?
: .
: : We appreciate that volume loss is the key and that narrow through wall pits can actually be missed by MFL, but in this instance the volume loss should have been detected.
: .
.



 
03:42 May-08-2002

haidar shahri

NDT Inspector
abadan refinery,
Iran,
Joined Apr 2002
3
Re: MFL - Tank Floors please explian MFL BEST REGARDS



 
04:14 May-08-2002

Mark Badrick

Inspection
Bahrain Petroleum Co,
Bahrain,
Joined Oct 1999
11
Re: MFL - Tank Floors As simply as I can: specific equipment uses Magnetic Flux Leakage (MFL) as a detection tool. Magnets introduce a magnetic flux into the material near to saturation level. Any localized reduction of thickness of the material will result in a "flux leakage" at the surface. A series of sensors on the equipment detect these leakage fields. However, the strength of the leakage field is a function of volume loss and is not a reliable indication of remaining wall thickness, which is why we have to UT proove up any indications found with MFL.


: please explian MFL BEST REGARDS
.



 
08:23 May-09-2002

John O'Brien

Consultant, -
Chevron ETC ,
USA,
Joined Jan 2000
278
Re: MFL - Tank Floors The simple answer is yes people still cut coupons for verification or take e.coupons with UT. There is a trend to go back to doing this more and more as a number of studies in the past two years have shown that MFL tank scanning is more operator dependent than was generally appreciated.

It is apparent that even with the same scanner results can vary between vendors and even between operators.

It has become apparent that there is a strong need to qualify operators and scanners by some form of performance demonstartion test and to conduct back up verification work.

API 653 is currently developing an Appendix which will give guidance on ways in which to establish performance demonstration. ASME 2001 Edition also has an MFL section although this was not primarily focussed at tank floor applications.

The other thing that people Miss is that although scanners may cover the whole floor there is a defect resolution below which they will not detect and you need to be aware of this as with some defects some as MIC they may be so small and isolated that they are below the resolving power of your system.

Think about the application, select appropriate procedures and qualifications and be aware of limitations - just like all other NDT methods.


 
01:54 May-18-2002

Rolf Diederichs

Director, Editor, Publisher, Internet, PHP MySQL
NDT.net,
Germany,
Joined Nov 1998
598
Re: MFL - Tank Floors I found an article which may help you:

Life Management Of Above Ground Atmospheric Storage Tanks
http://www.ndt.net/article/apcndt01/papers/935/935.htm

------------------------------------
: Up to a few years ago, Tank Floor Inspection at this refinery consisted of Visual, UT gauging, vacuum box testing and the cutting of coupons in selected locations. We have now taken advantage of the many service providers of MFL in the Gulf region. We have, however, still maintained the practice of cutting coupons and on several occasions have found discrepancies between what was reported using MFL and the physical evidence of the coupons.
.
: Firstly, are there any readers of this forum who use MFL and just rely on the UT proove-up to verify MFL findings, and secondly is there anybody that does the same as this refinery and cut coupons to verify the MFL?
.
: We appreciate that volume loss is the key and that narrow through wall pits can actually be missed by MFL, but in this instance the volume loss should have been detected.
.



 
04:18 May-20-2002

Michael Trinidad

Consultant, API 510 570 & 653
Marine Inspection Service Pty Ltd (MIS),
Australia,
Joined Jan 2003
138
Re: MFL - Tank Floors Unfortunately a lot of it comes down to operator error or lack of training. The MFL floorscanning generally has no or very little training and a lot of times operators are not aware of the circumstances which cause false signals.

Laminations, weld tags, surface roughness, arc welding in the vicinity are a few items that can give erroneos results. If the floor is rough then the signals will be questionable and this is also the case if it is very dirty.

Also some equipment is better than others. If you are going to do trials of operators and equipment use samples with fibreglass coating.


Kind Regards


Mike Trinidad


 


© NDT.net - The Web's Largest Portal of Nondestructive Testing (NDT) ISSN 1435-4934

Open Access Database, |Conference Proceedings| |Articles| |News| |Exhibition| |Forum| |Professional Network|