Flange Joint Live Tightening

[TXY]

Hi guys,

Looking for advice on performing live tightening on flange joint at 120degC. Prior to start up, it was torqued with the recommended 4 torque passes (30, 60, 100, 100% check). It has been some time since start up and the end user contacted to perform live tightening. My question is whether it is still necessary to perform 4 torque passes for live tightening. Or a clockwise check pass with 100% torque would suffice. Thanks in advance.

Cheers
XY

 

[LittleInch]

Doing anything to a live system has a significant increase in risk compared to doing anything to one which is not live.

Why does the end user think it need checking?

Is it leaking?

 

[NovaStark]

Hi guys,

Looking for advice on performing live tightening on flange joint at 120degC. Prior to start up, it was torqued with the recommended 4 torque passes (30, 60, 100, 100% check). It has been some time since start up and the end user contacted to perform live tightening. My question is whether it is still necessary to perform 4 torque passes for live tightening. Or a clockwise check pass with 100% torque would suffice. Thanks in advance.

Cheers
XY

Much like what LI says, if it isn't leaking, there wouldn't be any need to do an online tightening as that could impart more damage than benefit.

Is this a known problematic joint or operates in some condition where bolt relaxation is a sigifnicant issue ? 120 degC doesn't seem like a large enough temperature to cause it.

 

[TXY]

Thanks for the responses.

End user advises that it is leaking; although they don't seems overly anxious with this. I do not have any joint history as the initial start up tightening was performed by other bolting contractors. Information provided to me as of now is 24" 900# RTJ Flange connection with target bolt load 534kN (~27,000psi bolt stress). In my opinion, this is kind of low.

And back to my original question, in the event of live tightening, what would be the recommended procedure? ASME PCC-1 seems to be rather silent on this. My thoughts are 30% and 60% torque passes would likely not turn the nuts at all, so wondering if there are any merits of going thru the motions. And considering to minimize operators' exposure to this work, I'm inclined to propose that a direct 100% torque pass would suffice. Any thoughts?

 

[NovaStark]

If your bolts are B7 bolts, the yield is around 105 ksi at ambient conditions, so the 27 ksi is around 25% of the value.

Usually, we try to go to 50-60% depending on the situation.

With an RTJ gasket, you can try a slightly higher torque value however, note that the with torqueing of these gaskets, strain hardening can occur. So it should not be a repeated exercise.

 

[r6155]

Elongation control is the best practice, not torque. Checking elongation is easy.

 

[TGS4]

I'm with @r6155 here - elongation - likely by the turn-of-nut method - is going to be more robust in this situation. For torquing, you are assuming that you have the same nut factor as when the lubricant was new. It is not new any more. As a result, you won't know what your actual nut factor is.

Perhaps get a second opinion on the target bolt load. I know that it seems on the low side, but RTJs are different, and higher bolt loads can cause cracks in the flange at the bottom of the ring receptacle. Also, make sure that this is a PCC-1 Appendix O bolt stress, and not one from an ASME VIII-1 Appendix 2 calculation.

 

[LittleInch]

My thoughts are generally don't do it.

RTJ flanges are not the same as RF and tightening the bolts or even just checking they are all at their 100% torque ( noting the issue that 100% torque might no be the same as 100% of their intended stress due to rust, wear, lack of grease etc) may very well not cure the leak and if the leak is due to a crack or corrosion of the flange sealing surface you could easily increase the leak flow or lead to failure of the gasket ring.

In terms of bolting etc, PCC1 is for for new joints in the main. You may want to look at PCC-2 and the applicable section would seem to be article 311 on "hot bolting", noting the requirement to reduce to 25% of normal or max operating pressure.

Or use a flange sealing kit / company until you can de-pressurise the joint and actually see what the problem is.

What exactly is "some time" after installation, what does this pipe contain, what is it operating pressure and has anyone considered that tightening the bolts at 120C when already less stress is already present is a good thing when they cool down.

So in summary you need to think very carefully IMHO, if this operation will actually work or whether you will just be causing more damage or be back in a few weeks time, when the original fault just re occurs.

 

[r6155]

In this example, the nut rotation is not applied for verification: impossible.
Ultrasound is applied, compared to the original ultrasonic elongation report.

 

[TXY]

Hi guys, appreciate the discussion here so far.

Bolts are B16, so yes it is around 25% yield, which is why I feel it is kind of low for RTJ flanges. I agree that UT elongation might be a better way of verifying the actual bolt load, but I'm not too familiar with it. My thoughts are - If end user did not do this initially, we won't have the reference bolt length to calculate the delta and I don't think it is possible to 'unload' the bolts now.

I'm aware of PCC-2, but it seems to be applicable for hot bolting only (and not live tightening). My understanding is that these are two entirely different operations.

Regarding the lubricant, end user previously mentioned that it has dry up and is considering increasing the nut factor. Then again, how much of an increase is debatable.

Since the original post, they have not gotten back to me. It might be possible that they have decided on other alternatives internally.

 

[A8yssUK]

Late to the party again... but I am going to post anyway because this boils my blood!

You cannot find any information about live loading and/or hot bolting of bolted flanges from us gasket manufacturers because it is something YOU SHOULD 100% AVOID.
Here you are at 120°C and I assume with an #900 RTJ joint you have quite a high pressure (perhaps even a nasty chemical?)...so if the gasket and/or flange blows out with a PERSON stood next to it. I hope you as the engineer like jail time and if you're the installer....you need to look on the dark places of the internet for pictures and videos of what the remains of a person looks like after something bad happens.

"but the company! MONEY!... impossible!...blah blah blah..." You are an engineer not a marketing executive, do the job the right way.

Then the answer is easy, while not live, do the 4 pass again at the same torque. If the bolts don't move then you'll need to look into increasing the bolt load. (Either due to insufficient initial load or the dried up lubricant) TALK TO YOUR GASKET MANUFACTURER FIRST not a random forum on the internet.

Engineering is all about the details, which this thread is dangerously lacking.
RTJ gaskets work because they are a softer material than the flange materials. You want the RTJ to deform not the flange face.
So has anybody asked why is the initial recommended installation bolt load low ? an error ? or is it by design to prevent the flange faces cracking?
So what are the flange and gasket materials/types ? Perhaps it is a lower load designed RTJ with facings ? these also have a lower maximum load capability....

We also have no idea of the media or internal pressures, no idea if the end thrust has been accounted for.. no idea if any local regulations apply.. no idea about the initial calculations, no idea about any ancillary connection influences, no idea of the condition of the flange facings, no idea if the installer knows what a washer is and no idea how the bolt load was applied... so we really don't have anywhere enough info to give out any advise.

So TALK TO YOUR GASKET MANUFACTURER! They will ask all these questions for very good reasons. If they don't... then change your gasket manufacturer

 

[r6155]

Be care.
Bolted joints, like welded joints, must be performed by qualified personnel. Torque wrenches and other devices must be calibrated frequently.

 

[TXY]

@A8yss I don't understand why you had to make this so personal.

First of all, I'm not related to the end user and have no means to contact the gasket manufacturer, be involved with their internal discussions nor obtain any further information that could help assess the risk or the underlying issue.

And instead of venting your frustration on a 'random forum', have you considered writing in to ASME PCC committee to address the pressing concern that you had regarding live tightening? Wouldn't it be more meaningful to raise the issue to the right platform and get the (pretty much) most widely referenced guideline updated to let people know that they shouldn't even consider this as an option? The risk associated with live tightening has already been highlighted by many before you, just stating the obvious here doesn't help change anything nor make refinery maintenance work any safer.

By the way, if you have not noticed, PCC-1 guideline is written with representative of gasket manufacturers involved. The entire Appendix B only states to perform an engineering and risk analysis; it doesn't forbid the activity.

While I concur that you argument is valid; it is too idealistic. Imagine every engineer not getting a decent response from the gasket manufacturer asking for the manufacturer to be changed. Do you really think this is going to happen?

Anyway guys, there are no further follow up to this enquiry from the end user. I really appreciate the comments and sharing above.

 

[A8yssUK]

"why make it personal"

Its not personal for you just because your not the one that'll be turning the actual nuts? It's personal for them.

"First of all, I'm not related to the end user and have no means to contact the gasket manufacturer, be involved with their internal discussions nor obtain any further information that could help assess the risk or the underlying issue."

So your knowledgeable enough to know the bolt load appears low and your not going to do anything about it ?? Not even ask for more information or question it?. But your OK recommending that its fine to be tightened up to the non-live room temperature torque value that you have no idea why it is set to what it is?

"The entire Appendix B only states to perform an engineering and risk analysis; it doesn't forbid the activity."

"Perform an engineering analysis". Which I would say includes knowing what the flange and gasket materials actually are and their limits...

" Imagine every engineer not getting a decent response from the gasket manufacturer asking for the manufacturer to be changed. Do you really think this is going to happen? "

YES for god sakes YES stop using terrible gasket manufacturers/distributors, gaskets are literally by design the weakest link in a pressure system, why on hell would you use awful ones. Gaskets are like less than 1% of the total system costs.
Again if your gasket provider cannot answer simple engineering questions about the products they make, GET A NEW ONE!

 

[GD2]

There is no rule in any Code or RP that recommends bolt tightening in operating condition. All is done in atmospheric condition.
If the Owner still wants to perform at the operating condition, they should better do a "Risk Analysis", and get it approved by the Management.
There are many Owners, who will do it at the operating condition to stop a leak silently. As long as it is done with no incident, there will be no noise. If an incident happens, the Manager will initiate an investigation......
As an Engineer, you can't recommend this.