ASME Section VIII and Inlet Pressure Drop 3

[tkdhwjd]

Dear Colleagues,

Per ASME Section VIII, M-6, the inlet pressure losses will be based on the valve nameplate capacity corrected for the characteristics of the flowing fluid. Are there any exceptions to this requirement spelled out in ASME VIII? In other words, does ASME Section VIII also define specific casess/scenarios where owner engineer can take exception and use the actual relief requirement to calculate the relief device inlet pipe pressure drop?

Thank you in advance for your help.

 

[CJKruger]

The only mandatory requirement I think is in ASME 8 UG-135b1/f. And it does not mention the flow or specific criteria you should use.

So while almost everyone follows App M and API 520/521, you could probably deviate from it if a client insists on it.

 

[Chance17]

I have been involved with organizations that deviated as follows.
1) Fire - the contingency flow was used and not actual valve capacity. The reason - Fire is typically slow pressure generation and the valve is not likely to fully open
2)Liquids - the contingency flow was used and not actual valve capacity. I don't remember the rational
3)Allowable inlet loss limit was 5% instead of 3%. This was at a 60+ year old refinery. Revalidation of safety valves were finding a large fraction of deficiency over 3%. So they made a facility exemption to save money.

 

[Latexman]

When the pressure hits the huddling chamber of the disc and thrusts the valve open, how does it know to open just enough for the scenario at hand, instead of wide open like it's designed to do?

Good luck,
Latexman

 

[4Pipes]

Latexman is correct. Oversized valves / undersized inlet pipework has the same affect. Chatter is often a poor name for a very violent affect.
I was 10 yards away from an oversized PSV that went off like a very big jack hammer. Everything shook. A guy standing next to the valve stepped backwards in shock onto an unguarded platform edge. It was only his flapping arms that kept him on the platform. I came within an ace of needing a new pair of trousers. Neither of us was very impressed with the Process Engineer who spec'd the valve. He thought is was funny and that over sizing was not his problem.

 

[tkdhwjd]

Thank you all for your reply.
I concur with Latexman. What I am trying to figure out is whether or not deviating from this requirement would result in Code violation. If not, it will be much easier to get this approved by the site safety committee provided that a reason for deviation is based on sound engineering judgement/experiences.

 

[Latexman]

There is a reason the Code does not mention the flow or specific criteria you should use. Safety relief valves do not modulate (i.e. partially open or partially closed). The model is - they pop wide open at the set pressure, and go closed at the blowdown pressure. There is either zero flow or flow is at the relieving capacity.

ASME Section VIII UG-135 (b) says the flow characteristics of all pipe and fittings between a pressure vessel and its pressure relieving device shall be such that the pressure drop will not reduce the relieving capacity below that required, or adversely affect the proper operation of the pressure relieving device (i.e. chatter).

This aligns with what is in M-6, “inlet pressure losses will be based on the valve nameplate capacity corrected for the characteristics of the flowing fluid”. The “valve nameplate capacity (i.e. valve wide open) corrected for the characteristics of the flowing fluid” is the relieving capacity of the installation.

In both cases the Code is referring to the flow that occurs with a wide open safety relief valve and it’s pipe and fittings from the pressure vessel to the safety relief valve. There is no mention of using a sizing or required flow rate for the basis. There is mention that the relieving capacity (i.e. valve wide open) cannot fall below that required (i.e. sizing scenario). That means the upstream system must be designed with the safety relief valve wide open as the basis. I see no other way.

Chatter is the rapid opening and closing of the valve that will result in an unpredictable reduction in relief capacity, damages to the seating surfaces, and possible destruction of the safety valve. To avoid chatter, one must design the upstream system with the safety relief valve wide open, because that is what is really going to happen, and meet certain criteria.

IMO, substituting a flow rate from a sizing scenario is just folly, which may lead to future legal issues and liabilities.

Of course, the above is about compressible flow only; not liquids.


Good luck,
Latexman

 

[JAlton]

There are a few more things to consider. The nameplate capacity of an SRV is really only about 88% of its actual capacity. I don't say this to encourage taking credit for more than nameplate, but to caution that oversizing is easier to do than you might think.

Also, SRVs do not, as a rule, go to 100% lift (wide open) on initial pop. They typically go about 60% to 70% on initial pop unless it is a low set pressure where the spring is a little weaker. The full lift is only achieved at 107% to 110% overpressure. This is why One RIng Design, Process Safety-Relief Valves are not certified for ASME Sec. I Boiler Service. A typical SRV cannot meet 100% of relieving capacity by 103% overpressure as required by ASME Sec. I.

ASME Sec. VIII, Div. 1, Para. UG-134(f) is a recent revision which requires more attention to UG-135.

JAC

 

[Latexman]

All good things to know. We've used the actual capacity for designing downstream treatment systems and the nameplate/relieving/ASME capacity for the relief records for as long as I've been sizing reliefs, about 30 years. The dynamics of SRVs really don't come into play much for a relief sizer like me who looks at process upsets at 110% OP and fire cases at 121% OP, or I haven't figured out how to put that information to work yet. However, I can see it's critical for SRV designers though.

Good luck,
Latexman

 

[olcrazy1]

someone above said saftey valves do not modulate, this is not true, in pilot operated saftey valve does have a modulating design. And if your pressure drop is due to a long run of pipe before the valve inlet a great option is to go with a lipot valve becasue you can remote sense the inlet pressure directly to the pilot control. Also regarding chatter, it is a general rule of thumb that a valve is considered to be oversized if it does not flow at least 25% of its rated nameplate capacity.

 

[JAlton]

Olcrazy1 is right about Modulating Pilot Valves and remote sensing. Both are great ways to deal with Chatter due to oversizing. I understand Latexman's comment regarding direct spring response to pressure in the huddling chamber. The huddling chamber, invented in 1869, is still a good design for a PRV, but installation and operating scenarios are vastly more complex in 21st century industry than a 19th century steam locomotive. George Richardson was a locomotive engineer who wanted his Safety Valve to reseat quickly to avoid the loss of pressure and therefore slower speed until the pressure had recovered. He invented the huddling chamber to effect shorter blowdown and thereby designed the first successful spring loaded safety valve.

JAC