safety reilef valve

[Jasmine2000]

There is an SRV at the discharge of a heat exchanger the MAWP of this equipment is 270 psig and srv set pressure is set @ 270 psig. pump is pumping at 244 psig down stream of the HX. During the start up and shut down of the unit the srv leaked. The client I am working with he believes the srv will start to open at 90% of set pressure where I believe the valve starts to open at 98% ( 265 psig )of set pressure. He thinks pumping pressure is too close to the pressure where the SRV starts to open so it might damage valve seat. When exactly the valve starts to open?

The valve was sized to protect the EQ under fire conditon.

I looked at the inlet piping of SRV and I noticed is small by 1.5". Could this cause chattering and leakage? Thank you

 

[dcasto]

The valve should not open until 270 psig and be 100% open at 297 psig. You are closer to being correct on it opening. If you want some more room, switch to a pilot operated valve set at 297 psig.

 

[EGT01]

There is much we don't know about the system to which Jasmine2000 refers but if it is an ASME Section VIII pressure vessel protected by one relief device you are not allowed to have a set pressure greater than the MAWP.

As far as what pressure the valve will start to open, I would suggest you consult with the valve manufacturer as it relates to the type of relief valve and their recommendations. Relief valves will tend to simmer or leak prior to reaching the set point. Looking at two manufacturer's catalogs for spring loaded relief valves, both Farris and Consolidated recommend an operating pressure no more than about 90% of set pressure or, conversely, a set pressure 10% greater than operating pressure. A soft seat design will allow a closer operating margin but for an operating margin in the 98% of set pressure range, you will probably need a pilot operated type valve.

 

[davefitz]

Refer to the mfr catalog, or read the specific code requirements.

The code allows , or assumes, some inaccuracy in set pressure, as well as "accumulation", prior to complete lift of the srv. These innacuracies should be considered in the design of the process or the initial determination of the pressure part's design pressure.

 

[insult2injury]

You said that the SRV leaked during startup and shutdown, what about normal operation? It's possible that the transients that occur at startup/shutdown resulted in pressure spikes, but it really depends on your specific installation and procedures.

I2I

 

[maddocks]

If this is a conventional PSV, you can expect it to potentially weep at somewhere between 90 & 95% of set pressure. As dcasto pointed out, transition to a pilot operated RV and you can get to 95%. Expecting more from a spring based device is unlikely. You could also try a rupture disk or a compression pin assembly but you'll have to verify set point variations with the mfr.

 

[MJCronin]

jas...

Maddocks is on the right path with his suggestion of a rupture disc. They are permitted relief devices in accordance with the ASME (and other) pressure vessel codes.

They are useful where you want to approach, more closely, the necessary pressure setpoint of the device without the risk of leakage. As I recall, some discs can be made to fail at + or - 2% of rated pressure

You may also want to consider the combination of a ruptured disc and relief valve.

Another option, if indeed pump surge is your culprit may be to use some kind of surge tank ( with an internal bladder) on the pump discharge.

Star ?

-MJC

 

[BenThayer]

it sounds like you are having pressure fluctuations during start-up shutdown.

the rules of thumb we use is do not exceed 90% of set with a conventional PSV or it will start to weep. do not exceed 95% of set for a pilot operated valve.

and i would never consider a rupture disc in liquid service when i already have issues with pressure fluctuations. a very slight water hammer effect with valves cycling, flow surges, etc. will blow the disc.

 

[twr]

Jas,
The downstream pump is at 244 psig - does that become the inlet to the exchanger? When I think of a downstream pump, that is after the heat exchanger and what I would be looking for, is the pressure on the inlet/upstream side. Maybe I have just got the names mixed up.

We have a pressure and vacuum relief valve on an atmospheric silo, very low pressures, and that is definitely relieving prior to the maximum levels. The valve is rated to relieve fully at 3 psi - we fill the silo so that when the pressure is ~1.5-2.0 psi the valve relieves pressure. It stays at the higher pressure but won't continue to climb until we increase the fill rate.

The 1-1/2" size may not be a problem - you'll need to evaluate the pump capacity & flow rate compared against the manufacturer's specs.

Finally, check the discharge valves to see if they are open when you start up, and then gradually throttle back.

 

[MortenA]

look at API520 fig 1. It says simme 2% (typically). Furthermore the maximum expected operating pressure should be 10% lower than the MAWP.

Best regards

Morten

 

[psafety]

I prefer to stay away from rupture disks when there are pressure and temperature changes. That metal will cycle only so many times before it bursts.

 

[JoeWong88]

Jasmine,

When you said "...pump is pumping at 244 psig down stream of the HX. During the start up and shut down of the unit the srv leaked.", does the "unit" refer to pump ?

Presumably the unit is "pump", is there any shutdown valve on pump discharge close simultaneously at the pump discharge ?

is pump discharge to high point i.e. 5-10 m ?

What type of PSV ? Conventional ? Balanced ? Pilot operated ?

Is the check valve on pump discharge "non-slam" type ?

What type of pump ? Centrifugal ? reciprocating ?

If pump is reciprocating type, any dampener or surge suppressor on the pump discharge ?


I am sorry for asking so many questions but the information that you have furnished may not be sufficient to tackle the real problem i.e. PSV passing ?

If you read my question, you may quickly understood that i am suspecting pressure spike during pump shutdown. PSV on pump discharge passing during shutdown is pretty common if your system was not designed for it. Pressure spike could be due to

i) incorrect type check valve lead to valve slamming. Read "How to predict Check Valve Slam ?"

ii) long discharge line and discharge shutdown valve closing time is quicker than the pump "slowdown"

iii) liquid column hammering
.
.
.

Awaiting your response.