Temperature vs Brittle Fracture during Hydrotest 1

[Johnnyoo]

We are performing hydro tests outdoors in Alberta right now. I have been questioned about the effect on strength reduction when testing at lower temperature. I recall seeing a chart or graph where strength reduction was plotted against temperature. Can anyone assist here please??? We don't have an issue with our A333 Gr. 6 it is for hydro testing A106 Gr. B.
Thank you
Johnnyoo

 

[BigInch]

How low is low?
Keep A106 above -20F
You should have Charpy Testing results on file tested at the lowest of ambient or design temperature whenever below 0F

 

[MJCronin]

BigInch is correct. Most code rules only allow usage of simple carbon steel systems to -20F

You should know what piping code that you are building in accordance with and what the temperature limitations are for various materials

Most piping codes contain tables with allowable stress levels based on temperature. For A106 materials, ASME B31.1 and B31.3 have the same allowable stress at -20F as 70F.

MJCronin
Sr. Process Engineer

 

[LittleInch]

I an missing something here?. "hydrotest" implies water or glcol mix but even that will struggle below-20 c.

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

 

[BigInch]

He'll need pure glycol to do a hydro at lo-lo temp, so I'd guess he really means a pressure test with gas. Really not a good option when brittle fracture is possible.

 

[JStephen]

Maybe that means they're hydrotesting with water, but vessel/outdoor temps are ambient, so exact temperature of vessel would be unknown.

 

[TGS4]

50-50 water-glycol is good to -40°C/°F. The question is: how warm above the MDMT for a particular material should the hydrostatic test be conducted? For an MDMT of -29°C, and ambient temperatures hovering around -20°C our lower, can the OP hydrotest now or does he have to wait for a warmer day? And how much warmer? In his cross-post on the same topic, he indicated that the Code was B31.3.

 

[XL83NL]

Double-posted topic here. MDMT = -29 C.

 

[weldstan]

I would not recommend hydro-testing A-106 at -20F. While B31.3 permits design, based on thickness at -20F, the stress induced during hydro is considerably more than the design allowable. I have recorded ductile to brittle transitions between 20 F and 32 F for this material. Noted that the material is seamless; however, I have seen some flaws in seamless A-106 that would definitely have led to brittle fracture if toughness was compromised at low temp.

 

[weldstan]

If hoop stress induced during hydro-test is low enough as a percentage of yield strength, I would have no problem testing assuming that welds were properly made with -20 F filler metals and the stress imposed across the welds due to applied piping loads and the hydro-test were properly analyzed.

 

[davefitz]

The issue at low temperature is catastrophic failure of the vessel due to a pre-existing micro-crack growing at the speed of sound ( in the metal) due to complete lack of fracture toughness. The 2 fatal events that I know of are (a) a feedwater heater shell hydrotest in Dansville NY in the winter , shell failed and fell on a worker killing him and (b) the Florida steam drum failure circa 1970 when hdyrotested. Both events were winter hydrotests, and the testing company did not meet the ASME-required minimum hydro test temperature .

From the many failures ( and hundreds of worker deaths) of the North Sea oil platforms of the 1960's and 1970's it was found that every weld contains at least 2 pre-existing micro-cracks, from the initial weld arc and from breaking the final weld arc. The stress concentration factor at the edge of a crack is roughly 5, and during a hydrotest the average metal is close to yield stress. Insufficient fracture toughness combined with a crack edge stress on the order of 4-5 times yield stress is a formula for failure. In addition to that issue, is the separate issue that many welds are not stress relieved.

So the matter is not so much allowable stress as it is fracture mechanics.

"Whom the gods would destroy, they first make mad "

 

[davefitz]

As a follow-up on the topic of the historical background for using fracture mechanics to predict failure during a sub-zero hydrotest, see<

http://www.twi-global.com/technical...d-and-pipelines-the-past-present-and-future/>

How quickly we forget.

The BS British Standard codes of the 1970's incorporated fracture mechanics , which addressed the 1940's failures of the Liberty Ships and the 1960's failures of North Sea Oil platforms . Evidence of the BS fracture mechanics approach can be found in today's EN-12952-3 Annex B, comparison of weld notch factors, and the longstanding European explicit concern regarding weld fracture toughness ( not so explicitly a concern in US codes).

"Whom the gods would destroy, they first make mad "

 

[TGS4]

For ASME B31.3 piping, please note 345.2.2(c). However, that doesn't provide any specific guidance.

For ASME Section VIII, Division 1 pressure vessels, see UG-99(h), which says that

It is recommended that the metal temperature during the hydrostatic test be maintained at least 30°F (17°C) above the minimum design metal temperature, but need not exceed 120°F (48°C), to minimize the risk of brittle fracture.

For pneumatic tests, UG-100(c) makes the 30°F (17°C) recommendation a requirement.

 

[BigInch]

Possibly of interest,
Under specific environmental and stress conditions, methanol-induced intergranular cracking can occur in some pipe steel.

http://www.bst-tsb.gc.ca/eng/rapports-reports/pipeline/2013/p13h0013/p13h0013.asp