rejection rate

Date 07-19-2004 09:49

As a follow up to a previous massage I would appreciate info on acceptable rejection rates for the following scenarios:

Both cases are in a petrochemical environment using GTAW and SMAW, with rejection rates calculated as butt’s rejected vs. butt’s tested for pipe welding (butt rate). We weld mostly carbon, low alloys and stainless steels. Please note that we are using radiograph (X-ray) results only for our calculations. (We realize that in using radiograph results only we are not painting the complete picture, but this is the only inspection technique we have reliable data for)

1. What will be the benchmark for rejection rates, using radiographs only, for new fabrication under shop conditions using ASME B31.3 as inspection criteria for radiographs?

2. What will be the benchmark for rejection rates, using radiographs only, for maintenance welding under field conditions using ASME B31.3 as inspection criteria for radiographs? Where mostly old and new material are joined.

Thank you for your inputs.

Joe

By ajoy

Date 07-22-2004 11:26

When I was working in a petrochemical plant as a client inspector, we decided on an acceptable repair rate of 5 percent for welders who were deployed on the piping welds in the field. That is 1 in any successive 20 joints. Above which he would have to be rested.

The shop acceptance limit for the same should be less than this, lets say 4 percent.

This should be a good benchmark.

By HoutxInspect Date 08-02-2004 21:10

It might also depend on how bad the rejection is on any particular weld.

By KAJUN1

Date 08-03-2004 12:23

The repair rate of 5 percent is high to some standards for being job wide. The refinery I'm in welders get 2 repairs on one project, if more happens they leave for 90 days & retest to come back. Working a project now with nearly 400 welds that is 80% complete with NO repairs. Rod control is one of the best ways to keep repairs down.

By Engloid (*)

Date 08-12-2004 16:00

I'm not familiar with that particular code, but it seems almost like saying that it's fair to put 5% of the population in jail....reguardless of whether they commit a crime or not.

I'd have to look at each weld individually and decide...without concern of other welds or how many are on the job.

By Bill A

Date 08-12-2004 19:11

I think it is actually more like asking how many out of 100 people are statistically likely to commit a crime; it is not reasonable to assume any single individual is a criminal but it is reasonable to assume that out of 100 or 1000 randomly selected people there will be some criminals among them.
Welds are the same way. No welder can reasonably be expected to make an infinite number of welds without a rejectable flaw. Therefore,the question becomes, if a welder makes 100 welds, how many are likely to have rejectable flaws assuming the welder is of a typical skill level working under reasonable conditions. Once you have that answer (so far it looks like about 5% is a common response) then any experience that results in a significantly higher reject rate would suggest something might be out of whack and there is reason for closer study.

By CHGuilford (****)

Date 08-12-2004 21:46

To add to your comment....
When you think about why rejection rates are determined, it makes sense. Repairs cost a lot of time and money. Not just to do the actual work but also to document the reject, write the NCR, possible confer with the AI, recheck the joint, and so forth. Repairs can cost thousands of dollars. Shut down delays can cost hundreds of thousands. Naturally, you want good welders doing the work and someone, somewhere determines how many repairs are reasonable based on the type of work, skill of the available welders, and risk to be assumed.
It isn't really how many people will be assumed to make mistakes, it is how many mistakes can you tolerate from one person before you have to do something about it. On critical jobs, you want welders who have already done their practicing. Some rejection rates are very tough but I think most people are aware of what is expected of them when they take on the work. It isn't personal, it's just business.

Another way of looking at is, would you want a surgeon operating on you when he has a 5% mortality rate?

Chet Guilford

By Muskwa Date 08-28-2004 13:11

Rejection rate will always be a hot topic. I must say though in response to the question about "would I want a Surgeon operating on me if he has a 5% mortality rate?" that absolutely.......if the mortality rate without his attempt was 100 % and other Surgeons had a 50% mortality rate!

On any of the Union sites I have been on,a new construction/fabrication project that a welder had a repair/reject approaching 3% would be close to kicking his lunch kit down the road.

Site maintenance with old and contaminated parent material there needs to be allowances made for many factors. Generally speaking, a repair rate of 5 % would have a particular welder also explaining himself or "down the road".

But let us remember, that when a person speaks of repair rates.....they rarely mention other extraneous factors such as "if you do not give me 40 inches today.....you are fired" OR " a GOOD welder could do that" OR " are you done yet?" Bottom line, good welding is an art, and if a welder has never had a repair.......then he is not the welder giving you 40 inches a day ...day in and day out year after year.

Whatever repair rate you deem as acceptable on your job do two things in order to acheive it:
1) ensure that your welders are clear on your standards
2) support your welders when they ask for help ("hey boss, the wind is over 30km/hr....how about a hoarding for this TIG joint?")

"gets off the pulpit and gives others a chance"