specifying mill tolerance 3

[ARenko]

I've read a few posts in this forum that in the manufacturing of seamless pipe that one side of the pipe will be thin, while the opposite side will be thick (w/ respect to nominal wt). I understand that this thickness variation will happen, but is it true with respect to nominal? That implies that there will always be under tolerance except where the thickness is exactly nominal. Is that the case, or is it possible to find pipe that is say from nominal to +12.5% (i.e. no under tolerance)?

I ask because it has been suggested that we order our pipe in this manner (i.e. min wall = nominal) for the purpose of getting extra life in operation. My concernw are 1) added cost, although this may be made up for in extra life and 2) availability. Is it readily available or will we be limiting ourselves?

 

[ColonelSanders83]

1) Go get a firm quote from a seamless pipe manufacturer. This will more than likley end the argument.

2) Talk to your current seamless pipe vender. The 12.5% undertolerance has been the same value since the 1950's. You can ask if the vendor can live to tighter tolerances. I have heard on instaces where they could easily produce to 5% underthickness tolerance with the process improvements that have occured to date (this is somewhat dependant upon the pipe size and wall thickness), and this involved no greater expence as they were already producing the pipe in this fasion. Remember that the undertolerance is a maximum allowance, it is in the vendors best interest to stay the hell away from this value to avoid rejected pipe.

Just my two cents worth.

A question properly stated is a problem half solved.

Always remember, free advice is worth exactly what you pay for it!

http://www.ap-dynamics.ab.ca/

 

[rneill]

I'm sure the manufacturer can supply pipe to closer tolerances if that is what you want but this is likely to be more costly and have longer delivery since it will not likely be coming from stock.

Concur with the recommendation to get a quote from a supplier.

Note: if a manufacturer did have the ability to more closely regulate wall thickness, I suspect they would typically be targeting the thinnest wall they could while still falling within the tolerance band of the specification unless otherwise dictated by the customer. Logically, why would you want to supply more steel than you were required to ?

 

[MrBTU]

Our experience agrees with rneill's comment. We are a shell and tube heat exchanger manufacturer, and we slide tube bundles with baffles into our shells, which are pipe in most cases. What we have found over the years is that, as pipe manufacturers are able to control their wall thickness with more and more accuracy, they are making it closer and closer to that -12.5% wall tolerance. As rneill says, why would they want to give us more steel than they have to - and you can bet someone is getting a big bonus at the end of the year for the amount of steel they save...

 

[jte]

This is a good discussion. What is commonly known among piping and vessel designers and engineers is that the pipe wall under thickness tolerance is commonly 12.5%. But most folks don’t find out where this number comes from. Using SA-106 as an example, the 2007 edition lists this value in paragraph 16.3:

16.3 Thickness — The minimum wall thickness at any point shall not be more than 12.5% under the nominal wall thickness specified.

The answer to the question of “why don’t the pipe manufacturers target a pipe wall thickness all around the pipe of say 10%?” lies elsewhere. They cannot save 10% of the weight by doing this simply because they are also held to a weight tolerance – presumably to avoid just this type of thinking.

Weight — The weight of any length of pipe shall not vary more than 10% over and 3.5% under that specified. Unless otherwise agreed upon between the manufacturer and the purchaser, pipe in NPS 4 and smaller may be weighed in convenient lots; pipe larger than NPS 4 shall be weighed separately.

Interestingly, SA-53 is more tolerant of weights while maintaining the thickness tolerance: This specification allows plus or minus 10% - so in theory a fabricator might save a bit of steel by targeting a lower weight/thickness.

All the more reason to specify -106 over -53!

Alloy and Stainless will likely refer to SA-999 for general requirements and that spec has the same +10% -3.5% weight tolerance as SA-106. However, when ordered on a minimum wall basis, the requirements change – read the spec for details.

jt

 

[MrBTU]

Also, as manufacturing methods improve and tighter tolerances can be held in the process, perhaps the Code should revisit and tighten up the tolerances.

 

[jamesl]

jte,


I think the difference of weight tolerances between SA 106 B and SA-53 is due to the fact that SA-53 could be welded or seamless. The weight tolerance for SA 999 is for seamless pipe.

 

[rneill]

jte,

good point, I hadn't thought of the weight tolerance issue.

Also a good reminder that you should always check the applicable materials specification to confirm the wall thickness tolerance that is specified. Pipe grades such as A671 or A672 (made from plate) have significantly tighter tolerances than do the more common A106 materials. Also, line pipe grades such as API 5L material may have different tolerances as well. Specified tolerance may also vary by nominal diameter.

 

[Downhand]

I do not have access to ASTM A999. What is the minimum wall thickness for stainless under this code? Can specified thickness be determined from documentation such as the pipefitters handbook or is there a reference somewhere is ASTM codes that lists specified thicknesses?