Hydrotest pressure exceeded 1

[cespi82]

Hi Gents,

This is the case: 2" Pipeline SCH 80 API 5L Gr B. The system is rated at #600. The length of this pipeline is about 800 meters.

This system was undergoing a hydrotest as a part of the construction process. The test pressure was set at 2175 psi (company standard fro test pressure on gas pipelines at that rating).

Due to sudden ambient temperature increase, the pressure went above 3000 psi (3000 psi is the maximum reading of the gauge, so probably pressure could have gone higher).

Unfortunately the pressure test was unattended at that moment and test operator did not realize about this.

The fact is that the entire pipeline was exposed to a pressure much higher than the hydrotest pressure (over 3000 psi) and we are not sure if the pipes suffered damage due to the over pressure. There are no external evidences of damages like deformations or pipe movements. There wasn't any leak either.

Should we be concern about the mechanical integrity of this pipeline now after it has been exposed to this extreme pressure? Should we conduct any particular test to determine if it actually suffered some damages?

Thank you very much in advance for your replies.

Kind regards,

 

[racookpe1978]

Yes, it suffered damage. Where? We (and you) don't know.

 

[LittleInch]

My brief calc shows that the pressure to get to the SMYS, never mind actual yield for this size, grade and wall thickness (5.54 mm for Sch 80) of pipe, is 442bar or over 6,000 psig.

Therefore I suspect you've been lucky and you've not actually reached the yield stress of the pipe. If you have some idea of the temperature increase you can estimate the pressure rise - see this FAQ

http://www.eng-tips.com/faqs.cfm?fid=1339

I would just test it to the original pressure again. If you did yield it somewhere then that piece might be a bit thinner and would now be work hardened, but so long as it withstands your normal hydrotest then you should be Ok. Watch the pressure versus volume in graph very carefully and take lots of measurements. If the slope starts to tail off as you get to the test pressure then you are yielding somewhere and that point will be a weak spot or could fail in later years much sooner than otherwise.

Be very clear if there was only this pipe and no other connections or fittings.

I wouldn't use any of the flange bolts again....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Duwe6]

LittleInch has it exactly. Calc the stress and compare it to the pipe spec. If the specified mimimum yield isn't strong enough, recalc using the actual yield from the Material Test Reports [see why you have to save them?]. Only if if fails the second calc have you done something bad to your pipe.

And throw those studs away; nuts will be OK to reuse.

 

[BigInch]

Given L.I's calculation of 6000 psig to (not even) reach yield, and if you can ascertain that 2% strain limit was not exceeded (including at flange necks) and no leakage, flange deformation or other damage was found, why pressure test again. The code requirement is that you set and maintain the minimum test pressure for the minimum test time, which I venture to say that you apparently have. Codes do not prescribe maximum test pressures. Change out the studs and I'd suggest you are free to go.

I hate Windowz 8!!!!

 

[1gibson]

Well, OP doesn't specify if the duration of the test was met or not, so might have to re-hydro anyway. I would imagine it was pressured down as soon as someone saw the gauge maxed out.

 

[BigInch]

Correct. Since another thing that often happens is they don't notice that fact until somebody looks at the test chart. Anyway I left that possibility open with, "apparently", have meet test requirements.

If they did happen to meet minimum test time, what would you say about that?

I hate Windowz 8!!!!

 

[LittleInch]

I should perhaps say I based this on pipeline design code calculation not 31.3. It could do with checking as well...

I take the point about the fact that it might have passed the hydro, but would prefer to do it properly and make sure there was no excessive yielding anywhere. In this case going beyond the code minimum sounds like a good plan to me.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

Code minimum is test pressure and time. Code maximum is 2% strain. No damage found, so presumably it's somewhere in the middle, just where it's supposed to be, no?

I hate Windowz 8!!!!

 

[LittleInch]

Maybe, but until the OP works out what his maximum pressure went to I will advise differently. Just my opinion anyway.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

Of course, but it was you that said yourself (least in so many words) that it was highly doubtful he reached yield. I was drawing from your prior (possibly hasty?) conclusion.

I hate Windowz 8!!!!

 

[LittleInch]

I said I suspected he had got lucky, but until he tells us how much it went up in temperature we can't guess. I still think he's probably OK and second test is probably a bit OTT, but for the time and bother is probably worth it IMHO.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

If they did happen to not exceed yield pressure or temperature, what would you say then?

I hate Windowz 8!!!!

 

[LittleInch]

Write a report showing why the pressure rise was directly accountable to temperature rise like you would do for any other pressure rise in pipeline pressure test and sign it and submit for approval. The pipe might have leaked but any leak was maybe masked by the rise in pressure. As the pressure guage / transmitter maxed out, you can't show that the temperature rise was met by the pressure rise as you don't know what it was and therefore you need to do the test again - if nothing else to train the operator not to wander off as the sun came up....

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[cespi82]

Good morning gentlemen!!

I am gladly surprised about the quantity of answers and I do apologise for not replying back before. I went out on shift leave just few days after I posted this thread.

I just followed your advisements and conducted the Hydrotest again at the right test pressure, of course this time we ensured constant surveillance during the test. We have replaced the bolts/studs and the gaskets used during the previous hydrotest.

I was actually proposing the usage of a P-T chart recorder so that we will not miss any detail. However, this suggestion was not taken since the test period is short. I could not find any code or regulation to support me on this but I think it should be a must for the hydrotest. What would have happen in this case if test operators arrive before than me to the gauge location? they would probably have rapidly reduced the pressure so when I reach the place I wouldn't have noted anything. They would probably remain silent. Anyway, I would like to know your thoughts on this regard.

I went through the calculation path posted by LittleInch based on combined gas laws (Boyles, Charles and Gay -Lussac).

Since we don’t know exactly the temperature at the time of the hydro test, let’s just make assumptions:

Hydro test time = 10:00am – 12:30pm , 8th May 2014

Scenario 1 (conservative side)
Steel Temp at 10:00am = 35 C
Steel Temp at 12:30pm = 55 C
Initial pressure = 2175 psi at 10:30 am

Temperature difference = 20 C

20 x 63.84 = 1276 psi

New pressure = 2175 + 1276 = 3451 psi.


Scenario 2 (ultraconservative case)
Steel Temp at 10:00am = 25C
Steel Temp at 12:30pm = 60C

Initial pressure = 2175

Temperature difference = 35 C

35 x 63.84 = 2234 psi

New pressure = 2175 + 1276 = 4409 psi.

MAX pressure to get to the SMYS = 6410 psi

The two scenario shows that at given temperature differences, the pressure does not exceed the SMYS.
In order to exceed the SMYS, a temperature difference of 66.40 C is needed. (this is not realistic and attainable in ambient condition)

LittleInch, is this the calculation path followed by you?

Thank you very much all for your replies and once again my apologise for not answering this before.

Kind regards,

Cesar

 

[Duwe6]

A chart recorder is not mandatory. However, on a big hydro like yours, I would have an inspector 'ride the gauge' and keep a time-marked written log of the pressure. You would not have had the overpressure if the gauge and a vent valve were manned. And a pressure rise it typical and almost unavoidable with a properly vented line. Either man the gauge or use an adjustable PRV.

 

[LittleInch]

Looks about right to me. I used the FAQ and also another code which calculates it a different way and for the 35 degree increase case came to a change of around 2,400psi.

All the codes normally say is that a record of temp, pressure needs to be taken with start / finish as the minimum, but every 30 to 45 mins is recommended. Chart recorders are normally just seen as good practice or used to prove to an inspector what happened.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way