Low pressure vessel hydro test 1

[Shaheryar]

Is there a code which states that pressure vessel with design pressure of 15 psig needs to be hydro tested at minimum of 50 psig.

 

[metengr]

None that I am aware of here in the US. Pressure vessels at or below 15 psig design pressure are normally exempt from ASME Boiler and Pressure Vessel Code, Section VIII and do not need to be hydrotested and stamped.

 

[Shaheryar]

Thanks metengr, I am aware of the clause which exempts vessels with less than 15 psig from ASME section VIII.
But what if the vessel is designed for 16 psig, would it have to be hydro-tested to anything greater than 1.3 times i.e. 50 psig

 

[metengr]

No.

 

[codeeng]

Taking into account any surplus thickness and any corrosion allowance you could easily apply a test pressure several times the design pressure (commonly know as a new and cold test). As long as your limiting component is <90% yield then you're ok.

 

[metengr]

What practical purpose would it serve to perform a hydrotest above 1.3 to 1.5 X design pressure of 16 psig regardless of corrosion allowance or subjecting the vessel to 90% of yield? Subjecting the vessel to 50 psig makes no sense and serves no useful purpose.

 

[codeeng]

Firstly you should read-up on UG-99 which clearly states the different types of tests including new and cold. In my world a N&C test is more popular that a 1.3DP test. The theory behind a N&C test is that it produces a stress equal to or higher that a future field test (corroded).

There are many applications where an end user specifies 15 psi (or less) so that he does not go through the expense of registration, AI inspection and future in-service inspections. He may still want some additional mechanical integrity and designs for a higher pressure. To support this additional integrity he specifies a higher test pressure.

 

[metengr]

Actually, I am very familiar with UG-99. Conducting hydrostatic testing above 1.5X DP but remaining below 90% of yield does nothing to assure added mechanical integrity, other than you have no gross defects and the material will not yield in service. If the component is designed in accordance with code rules, you would not develop macro-yielding until you reached a membrane stress that is above 90% of the yield strength of the material. The only benefit of higher hydrostatic test pressure is to assure no defects in the material or at fittings and to provide local or micro-yielding of material at local transitions or stress concentrations, which actually reduces susceptibility to crack initiation, this is about it.

Too much emphasis is placed on hydrostatic testing of in-service components to assure a false sense of security that by going to higher test pressures you have additional mechanical integrity. Also, more importantly, the effect of elevated service temperature is not included with hydrostatic testing.

 

[codeeng]

Well at least I attempted to answer the original question and I can assure you that N&C testing is real and does exist. By the way 2 main definitions of mechanical integrity would be no defects and no yielding so I'm not sure where you're coming from?

 

[metengr]

Well then, your definition of mechanical integrity is the same as mine – no leaks. Have a great day.

 

[Katylee]

Sorry metengr but I agree with codeeng. In our refinery we specify new and cold testing for all new construction. A rough example using the original 15 psig design:

Calculated thickness: 0.2"
Corrosion allowance: 0.2"
Since we can't get 0.4" thickness we round up to 0.5"

Test pressure = 0.5/0.2 x 15 x 1.3 = 49 or 50 psig

It could be even higher if the design temp was high since the stress ratio would be greater than 1.0.

 

[metengr]

Katylee and codeeng;
That is fine to agree with one another, with the example provided I see your line of reasoning for wanting to increase the test pressure for new vessel construction. However, in my world, this approach is not necessary and would not provide any benefit in assuring further design integrity beyond what the design margins in the code inherently provide, and certainly would not increase the ability to find leaks from gross defects. A 1.5X DP hydrostatic test is plenty of pressure to find leaks.

For in-service vessels, hydrostatic testing of aged equipment must be done with extreme caution, and in some cases I would not recommend it. If anything, this could result in serious problems related to service-exposed materials because they may actually have reduced fracture toughness. Any future benefit of attempting to use a higher pressure hydrostatic test goes away as soon as the material is placed into service.

 

[Katylee]

I should have also compared this same example using the 1.3DP basis:

Shop stress during test = 0.2/0.5 x 1.3 = 52%

Operating stress (corroded) = 0.2/0.3 = 67%

This is a very undesirable situation and thats why we always insist on new and cold testing.