Hello guys,
I agree totally with Al.
In Australia some testing authorities have it clearly stated on the test report (be it destructive or non destructive) that the welding was not witnessed by the testing authority. In some states in Australia it is all self-regulatory and that is where the scary part comes in.
1 One welder welds 15 test coupons with no third party witness. An NDT request with 15 different guys names on and all done in the 6G position (in reality they could all have been rolled on the bench) and they are sent off to the local laboratory.They all pass and next thing you have 15 welder qualification certificates being issued to 14 guys who have never even done a test.
2 7 welders are employed on a piping job with 10% RT requirement and they have completed 100 welds. Instead of performing 10% of each welders work it is 10% of the lot so 10 welds of the best welder are shot and none of the other 6 welders.
3 Similar scenario and all 10 welds that are shot just happen to be on weldneck flanges.
These are not hypotheticals, these kind of "shenanigans" go on every day.
You may well say how do they get away with it, easy.
Take for example an Alumina Refinery, what they do well is make alumina, if they don't have a CWI employed they may have nobody with a welding background. So they rely on the "honesty" of the "supposed" professional pipe welding outfit to give them a quality product.
Pipes passed hydro, 10% " random " RT has been done and everyones happy. Does the Refinery have a clue that they may have been duped, No.
Excuse me for sounding so pessimitic but I have been around a fair while both on the tools and as an inspector and I have seen stuff that would scare the daylights out of you.
And it is only when someone is seriously injured or killed that anything is done.
Self-regulatory - what a joke.
Sorry for the rant, hope everyone has a great weekend.
Shane

In New Zealand it is totally opposite, it is an unwritten law that all ASME IX welders quals (which is the predominant code) are to be witnessed by a third party. When tender documents are submitted if the welder quals were not witnessed by a third party they generally go to the bottom of the pile.It increases the costs but it is generally done across the board so that all the costs are transparent and everybodies tenders reflect the same costs. It also helps to raise/maintain quality standards right across the industry.

Hello once again.

The thread is taking on some interesting twists.

I agree that most welding standards do not have a requirement for a test witness. However, the testing must be done under some sort of supervision and hopefully that supervision is by someone knowledgeable of welding.

I'll give you a couple of examples where I've seen the testing/qualification of welder go astray. Company XYZ tested their welders by handing the test coupons to the welder, instructing them to weld several samples and hand in the "best" for destructive testing. The certification papers were completed indicating the welder tested in the vertical, horizontal, and overhead positions, when in fact all the test plates were welded in the flat. I know this to be a fact because I was new employee and I watch the welders weld the plates. Another company I worked with for a short while would give the welders the test pipe and tell them to weld them up, and once again, hand in the best sample. The pipe coupons were welded on the bench and rolled as the welder completed the groove. The test reports listed the test position as 6G.

A different situation arises when qualifying the welding procedure. Again there is no requirement for a test witness. Company BBB submitted several WPSs and supporting PQR when my client and I visited their fabrication facility. After looking at the PQRs I asked who the test witness was and they replied that there was no requirement for the test to be witnessed. I agreed, but I noted that there were entries for travel speed, voltage, amperage, etc. The "welding engineer" replied that the welder provided that information. I made the comment that this was a good welder and that I would like to meet him. The "welding engineer" said that would not be a problem and off we went.

I was introduced to the welder and I asked him if he remembered qualifying the PQR in question. He replied that he qualified not only that PQR, but all the PQRs.

I asked him if he recorded the welding parameters. He replied that it was clear to him that I was an idiot that knew nothing of welding or I would know that it was impossible to deposit weld metal (using a manual process by the way) and watch the meters on the welder and operate a stop watch simultaniously. He made my point. I asked where the test data came from and his reply was that it "beat the s**t out of him, because it didn't come from him!"

The welding parameters and all other data recorded on the PQR are the actual values recorded and used during the welding of the test pieces. I call data entered after the fact "pencil whipped", it is meaningless and defeats the purpose of the PQR. After all, the purpose of the PQR is to record all the actual (real) information and test data that resulted in a test piece that passed all the nondestructive and destructive tests. Pencil whipping the welding parameters is falsifying the test information.

I've had several instances where the information was clearly falsified and I have questioned whether the tests were ever performed. I had PQRs that were welded and tested overnight. It is highly unlikely that a fabricator has the capability to weld, perform NDT, cut, machine, test samples, and fill in the paperwork in one or two days. I'm not saying it is impossible, by highly improbable.

Worst yet, assuming the testing was performed as per the welding code or standard, what do you do when the information is simply wrong? Do you, the reviewer, assume the mistakes are simply typographical errors. At what point do you make the determination that the individual tasked with the paperwork isn't competent and when do you raise the question, "Were the test plates or pipe actually welded using the parameters recorded and were the destructive tests performed in-house done properly?" I have a situation as I type that involves several PQRs and three WPSs. The number or errors on the paper work fills several pages. Are the errors simply typos? Can I believe the test samples were tested using the proper bend radius? Can I believe an individual that can't fill in the paperwork properly interpreted the code requirements properly? Can you have short circuiting transfer at 220 amps, 28 volts, using 0.035 inch diameter carbon steel electrode? Each time I ask for correction, new mistakes are noted.

So, do you have to have a test witness? You tell me. What are you going to do to provide a level of assurance to a jaded and experienced client that the tests were performed as indicated by your paperwork? The third party test witness lends credibility to your welding program. Does the third party need to be a CWI or SCWI? Again the answer may be "no", but I hope the individual understands what is required.

I guess the bottom line is this, if you understand the code requirements, if you understand how to complete the necessary forms, if you supervise the testing, if you or someone that is competent record the actual test data/welding parameters, a third party witness is not required. If that was the case and if every company did the right thing by providing the necessary training so that someone within the organization had the necessary competency, I would be doing something else for a living.

Good luck - Al