Monitoring a leak through pipe crack 2

[NovaStark]

Good day all,

Currently on my facility there is a steam line that operates around 600 psig and 700F which has a 2 inch long circ. crack on it.

So its obviously a safety hazard and there is a plan to repair soon. However until that time I was mandated to take noise readings as a measure to show that the leakage rate is not increasing.

My question is essentially based on anyone's experience what is the nest way to monitor a leak ? (I personally don't understand the logic of just taking an overall dB measurement and saying that louder = more leaking when I think the flow dynamics and the crack geometry, material and so on are too complex for something that simple - you can correct me if I am wrong in my thinking)

 

[europipe]

I think you are right, nobody can predict the busting time on basis of the decibels rise (or not).

 

[LittleInch]

Has no one heard of these:

" a steam line that operates around 600 psig and 700F which has a 2 inch long circ. crack on it"

Failure of a line like this would almost certainly lead to injury, maiming or death. This appears on the face of it to gross negligence on the part of the management team.

To allow it to continue to leak is unheard of in the modern world. Is there no safety department there?, no government inspector or authority?

Noise is no indication of mass, only velocity as it emerges. As the crack increases then it might get "louder" or quieter.

There is no best way to monitor a leak, only a plan to fix it without delay. Leaks from glands and flanges / mechanical joints might be tolerated, but "cracks" are inherently unstable and do not grow at a steady rate.

I realize you're in a poor position to insist on anything, but I would still put your concerns in writing to someone and print and keep a copy.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[NovaStark]

Ironically LI, my safety department seems fixated that measuring noise is a good indicator of the size of a leak. I've asked several times to see any study or science behind that claim but the same mantra of "bigger leak, more noise" keeps coming up.

(We do use clamps though in case you though!)

But I suspect that soon we will fix the problem once someone higher than the upper management understands the implication of the worst case scenario of overpressurization or occasional loads.

 

[weldstan]

LI
A star for you. And a jail mark for Nova Star's Safety gurus.

 

[NovaStark]

@weldstan

More than fair haha

 

[NBrink]

Have seen tables to estimate leak rate based on plume size, but even that can be greatly affected by temperature and relative humidity. Was part of an effort at one time to come up with criteria on when to repair steam leaks based on utility costs (Only things like packing leaks in areas where there was no safety concerns).

But yeah, leak that big, fix it. And hard barricade a fairly large space around it until you can get it isolated/repaired.

Nathan Brink

 

[LittleInch]

So why no clamp now??, not some vague time in the future, that's what I don't understand.

For something like the old "we can't shut it down" line is simply not relevant.

Can you get us a photo ( from a safe distance) ?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[NovaStark]

It is up in a pipe rack so its extremely congested for someone to install a clamp (I've already told the people in charge that I am not sending anyone under me up there do anything).

They're basically trying to keep the facility online until a predetermined shutdown next week.

 

[BigInch]

Is there no safety and inspection authority in your country?
You need to give them a message. I will call them for you, if you don't feel it is appropriate for you to do so.

 

[jte]

From a safety perspective, yes this could cause damage - so a high consequence. You can mitigate risk, a function of consequence and probability, by reducing the probability. In this case, barricading a substantial area to limit personnel exposure. The consequence of a "guillotine" failure and subsequent whipping action of the pipe on nearby piping and equipment should be included in the risk assessment. I've been involved with pipes which opened up and wound up bent to beyond 90 degrees.

The idea of a bolted box (clamp) type temporary leak repair is ok, but not without due consideration for the mode of failure. In this case, the OP NovaStark has identified it as a circumferential crack. With an unsound (sorry, couldn't resist!) means of monitoring crack propagation, putting a simple bolted box on would be insufficient. One must either design the connection between the box and pipe to transmit the longitudinal load, or install separate strongback devices to restrain the pipe from pulling out from the box longitudinally. That's a long way of saying "Read ASME PCC-2 Article 3.6 with a focus on paragraph 2.2" In that article, you'll see words like Therefore, leak clamps shall not be used when cracks are present, unless... That's a "shall" not a "should" statement for a pretty good reason.

 

[NovaStark]

There was talk of performing a level 3 API 579 assessment on the crack to determine if it could be deemed "safe" until a week (assuming no crack growth...somehow) and talks of clamping it.

But clearer heads prevailed and the system was shut down to repair it this evening. But I really wanted to know if that noise thing was actually done in other parts of the world, I thought I was going insane going against a well known practice. But just for information purposes, the cracked pipe is on a Tee with the branch having the crack. The branch is connected to an RV which lifted and soon after the crack was found. From that information no one wanted to keep the system online.

 

[LittleInch]

V glad it's now being dealt with correctly and from here I'm sure you got the idea that the noise idea was not the way it's done.

Also well done for refusing to send anyone up they to fix it while it was live. That's always been one of my red lines. If I wasn't prepared to go and do it I wasn't going to send someone else.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[NovaStark]

Yeah thankfully it'll be more reliable now.

Also reading up on API 579, does that actually hold for through wall cracks ? As in how does a code deem a through wall crack to be acceptable ?

 

[NBrink]

There are assessment procedures for through wall cracks. Have seen them used to justify a fillet welded patch as a temporary repair by confirming there would be no unstable propogation...

But do note that 579 isn't a code per se, it's a recommended practice.

Nathan Brink

 

[BigInch]

Clamps are inherently better suited to repair pinholes.

 

[DekDee]

I worked on a huge lattice work crane boom where the Japanese manufacturers user manual recommended checking for cracks by hitting with a hammer and listening for the differing sounds.
Either something went totally wrong in the translation from Japanese to English or there are some very weird inspection techniques out there.

 

[LittleInch]

DekDee - No - That's been around since the early days of steam engines - There was a specific person called the wheel tapper who would go round tapping all the cast steel wheels and listening to the noise. This would detect cracks in the wheels before it disintegrated.

Apparently still done in many countries

https://en.wikipedia.org/wiki/Wheeltapper

http://www.railwayarchive.org.uk/stories/pages.php?enum=LE124&maxp=16&pnum=7

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[DekDee]

I think VT combined with MT may be a more professional way to perform inspections.
Can just picture this:
"Shane, how did your inspection of our 500 ton crane boom go - acceptable or not ? "
"Sounds good boss ! "

 

[LittleInch]

But probably not as effective!

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.