Stub-In vs Weld-o-Let?

[UW1981]

We currently allow o-lets when the branch is 2" and smaller (limited by 1/2 size of carrier pipe). We then require stub-ins w/reinfocing pads in the "intermediate" ratios (ie, hard to find reducing T's), and then reducing T's on the higher ratios (which are commonly available).

The problem is that there are some applications where a stub-in can be made without reinforcing pad...which means we probably could use an o-let. The "conventional wisdom" around here is that the stub-in is a better connection since it is full penetrating as compared to the o-let (even when the pad is not required)

Is there any information comparing stub-in connections with o-lets regarding strength?

 

[1969grad]

This discusion will be interesting!

In my opinion, I would use an o'let over a bare (no pad) stub-in as I think the joint would be stronger. Granted there are some precautions about the header weld for o'lets which are documented on the Bonney Forge (or whatever their name is now) site.

I say stronger because there is more metal in the reinforcement zone. And, Bonney Forge does have some tests to verify that position.

Like you, use of o'lets for NPS 2" and less is normally used and is commonly permitted in various company specs.

For connections larger than 2" for which reducing tees are not available then the choice usually permitted is a stub-in with weld saddle as a minimum or a reducing tee with a reducer on the branch outlet.

 

[NozzleTwister]

What is the commodity, temperature and pressure?

I can't rememeber ever seeing, or even seeing allowed in a pipe spec, a no-pad stub-in for a branch 2" NPS or below in hydrocarbon service.

Installed correctly, the o-let should give you a stronger connection as well as provide integral reinforcement. Aside from that, your SIF's (Stress Intensification Factors) at the branch are going to be about double for the unreinforced stub-in as they will be for the o-let. That's a pretty hefty penalty if you have any temperature, thermal cycling or vibration.

Even for non critical service, a robust connection would be desired if there is any possibility of abuse like being bumped, stepped on, etc.

NozzleTwister
Houston, Texas

 

[UW1981]

I am trying to remember a specific application...but let's say it is a sch 80 2" off of a sch 40 8" line - all seemless A106 piping. Line is operating at 500 psig and 90 F...raw natural gas with about 1% H2S. Non-cyclic. Now...I have not done the calcs on this particular configuration, but I think this situation will not require a re-inforcement pad according to code.

 

[NozzleTwister]

Depending on your corrosion allowance, you may indeed, not need reinforcement to satisfy the code. But as I touched on before, that's not your only consideration.

I can't speak for your specific situation, but I think in general, non-reinforced stub-ins for small branches is not a good idea.

In addition to the usuall o-let, many operators today require the use of Sch. 160 nipples between the branch and the first block valve. A few pennies spent to get a little beefier design is much cheaper than fixing a leak.

Just my opinion

NozzleTwister
Houston, Texas

 

[massiveT]

Reinforcement is often not required for pressure reasons. However, piping stress on you branch connection will often require reinforcement. I don't know if you done a stress analysis on this but it is always a good idea to reinforce at you branch connections.

 

[UW1981]

Folks, thanks for the feedback.

I talked to the Engineer at Bonney Forge...he had no information comparing stub-ins (reinforced or otherwise) to o-lets. Perhaps I was talking to the wrong engineer....but if there is a contact you can share, it would be appreciated.

Our specs are not explicit regarding the re-inforcing pad on a stub-in - we need to follow code, but some folks have interpreted our specs to use re-inforcement only when required. But I think we need to move in the direction of having at least minimal re-inforcement for any stub-in based on the comments above (even if the calcs show we do not need a re-inforcing pad).

I guess another issue...kind of relating back to the original question (sort of)...what is the size limit for using o-lets. Again, we limit their use to less than 2" branch connections. However, we have folks (internal and vendors) that would love to use them in much larger sizes...apparently some folks believe there is some level of inherent re-inforcement in 0-lets. Comments?

 

[NozzleTwister]

UW1981,

I'm not surprised Bonney Forge couldn't provide information that might reduce their sales. Also, there is a liability issue. Reinforcement is required many times for reasons other than area replacement at a branch. Area replacement is straight forward. Other requirements for reinforcement are up to the design engineer.

In the petrochem industry that I'm familiar with, the break between butt-weld piping and socket-weld (or threaded) piping is generally at 2" NPS. 2" NPS & below is socket-weld and above 2" NPS is butt-weld. I've seen this break at 1 1/2" NPS but it's not that common. Assuming the break is at 2" NPS, then all of your piping and branches for run sizes 2" NPS and below would have socket-weld or threaded full or reducing tees.

When your run size is above 2" NPS, all of the branches 2" NPS and below would be sock-o-lets or thread-o-lets. For piping above 2" NPS, welding tees are used for size on size connections and reducing welding tees are common when the branch is one line size smaller than the run.

For the rest of the branches for run piping above 2" NPS, depending on your reinforcement requirements, you can have a stub-in, stub-in with re-pad or weld-o-let. The guideline here is usually economics or client preference. If you don't need reinforcement, go with a stub-in. You can always add reinforcement to a specific branch if required to reduce the SIF's.

Where reinforcement is required and based on economics, weld-o-lets are used up to 6", 8" or 10" branch sizes and re-pads are used for the larger braches. This is based on the total installed cost comparison between using weld-o-lets and re-pads considering material costs (weld-o-let vs. pad), the amount of welding required and labor. Naturally this changes year to year and location to location.

Some of our clients have preferred to use weld-o-lets all the way up to 20" NPS (that's a chunk of steel) and some clients have wanted all connections reinforced, whether required or not.

Back to your original question, my experience is that all branches 2" NPS and smaller are reinforced with an o-let of some kind.

I hope this helps. If you're looking for something in black and white, you're not going to find it. The guidelines I've given are general and based on my experience in my industry, of course there are exceptions.

Good luck,

NozzleTwister
Houston, Texas