PVs exposed to temporary sudden increases in pressure

[NovaStark]

Typically a PV will be designed to a fixed design pressure and temperature which is used to determine thicknesses and so on.

However is there a way to quantify what effect a sudden increase in pressure (still below design) would have on a vessel?

For example, say a vessel is designed to 700 psig at 200F and operates at 400 psig at 150F. It is temporarily exposed to a pressure change of 600 psig/hr for 1 hr at 150F.

Is it as simple as saying 600 psig/hr * 1 hr = 600 psig and since all the components are designed to 700 psig, no rupture should occur ?

Or is it like impact loading where the velocity of an object falling will increase the static stress by some value influenced by velocity ?

I haven't really seen anything in VIII about temperature or pressure excursions but in B31.3 there is a provision for temperature excursions for specified durations.

 

[TGS4]

There's no equivalent "momentum" behind a sudden increase in pressure. Therefore you analogy with impact loading doesn't work.

Provided that the pressure is less than design and the temperature is less than design temperature, then you're good to go.

HOWEVER, if this is a repeated occurrence, then you would need to consider the effects of the resultant fatigue.

 

[LittleInch]

"Is it as simple as saying 600 psig/hr * 1 hr = 600 psig and since all the components are designed to 700 psig, no rupture should occur ?"
Yes. In this case the time is irrelevant. 1 second or one year - makes no difference.

There is clearly an increase in stress on the PV, but this is less than the design P & T and therefore is satisfactory.

The issue with very rapid changes in pressure is more that this often spikes the pressure above the DP. If the pressure simply rises very rapidly, you might get a reaction force from the piping affecting the vessel, but not the vessel itself.

It's not much different from saying you have a 700W motor which normally runs at 400W, but then has a power increase to 600W - What's the problem?

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

 

[NovaStark]

LI so then in the case of a pressure spike then or let's say a rapid pressure increase above design for a short duration (1 hr or so), how does a reaction force arise if according to RGS4 there is no momentum change occuring ?

Or is it that whatever the psi/hr value is you use that psi and recheck how much stress a shell may experience ?

(By the way I am just asking for knowledge purposes as I am aware that a relief valve should be set below or at your design pressure)

 

[LittleInch]

You don't get a force on the vessel, but you can get loading from the attached pipework due to end cap forces / reaction forces on any bends outside the PV.

Pressure vessels have no design allowance above their MAWP unlike pipework so lets not go there.

Time has no impact and is irrelevant in what you're describing.

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

 

[NovaStark]

I understand what you meant now.

Much thanks for the insights LI and TGS4.

 

[1gibson]

Change of 600psi/hr is 10psi/minute or .167 psi/second, in what context is that sudden? Something like water hammer or a pressure surge would be more along the lines of 100 psi spike in a fraction of a second, then return to previous pressure, but you clearly aren't asking about that.

 

[NovaStark]

Change of 600psi/hr is 10psi/minute or .167 psi/second, in what context is that sudden? Something like water hammer or a pressure surge would be more along the lines of 100 psi spike in a fraction of a second, then return to previous pressure, but you clearly aren't asking about that.

Would be in the case of a process upset which there would be a basic spike (wherein the pressure transmitter would read pressure in a rate of change rather than a tpyical pressure reading). However I went for the case where I thought it would be "worst" case to ask i.e. 600 psi/hr * 1hr = 600 psi as opposed to to 600 psi/hr for 0.1 min to give 1 psi. Based on the responses it appears that the rate doesn't make a difference as only the pressure componenent is important.

 

[georgeverghese]

There are some guidelines in the ASME BPVC on the criteria for which fatigue analysis is recommended, and this is based on the frequency and rate of rise of pressure cycles expected over the lifetime of the vessel, even when the pressure cycles do not breach the design pressure - a pressure vessel engineer may advise where to find this.

 

[TGS4]

There are several screening methods in ASME Section VIII, Division 2, paragraph 5.5.2.

 

[georgeverghese]

A sharp pressure rise can also mean feed flow is well in excess of the process design flow - this may, in addition to metal fatigue on the pressure containing members, also cause excessive dp on some internal devices in the vessel ( feed nozzle, trays, demister pads etc), which can deform these devices. So a process design check on the effect of these high feed flows would also be necessary.

 

[NovaStark]

There are several screening methods in ASME Section VIII, Division 2, paragraph 5.5.2.

But would there be merit in spending time/money in performing a fatigue analysis for something that happens once in every 15 years ?

A sharp pressure rise can also mean feed flow is well in excess of the process design flow - this may, in addition to metal fatigue on the pressure containing members, also cause excessive dp on some internal devices in the vessel ( feed nozzle, trays, demister pads etc), which can deform these devices. So a process design check on the effect of these high feed flows would also be necessary.

This is also what I was thinking as well with my impact loading analogy. But like you said this would be limited to more of a process design case, and mechanically speaking, the shell and heads and internal piping or whatever would be affected how LI and TGS4 said.

However sometimes a problem with the process design check is that there aren't enough gauges or transmitters to tell you what you'd need to know to quantify these things as I've learned!

 

[TGS4]

A simple screening can take a skilled engineer all of 5 minutes. A more complicated could stretch into an hour. It's not a fatigue analysis, but a screening for exemption from a fatigue analysis. Quite simple. Really.