Cyclic discharge of pressure relief valve 1

[poetix99]

I have a Pressure Safety Relief Valve (PSRV) in a steam line supplying some in-house test equipment. The PSRV is intended to protect the test equipment from over-pressurization.

Nominal crack/full-open pressures are 330/365 psig. During occasional trip-outs (unplanned, sudden shutdown of the test equipment), fast closing valves will isolate a section of the steam piping, including PSRV. The closure sequence results in that pipe section being over-pressurized, and the PSRV will discharge, as designed.

The discharge from the PSRV is rapidly cyclic -- the valve audibly opens and closes until the piping pressure is <330 psig again. There are perhaps tens of cycles per event(20-50?) at a frequency of approximately 5Hz. These are hard to quantify, as we've made no recording of any sort, but should be adequate for this "conversation". The frequency of trip-outs is also hard to characterize, as that is never an intended event, but it would NEVER exceed 50 trip events/year. Therefore I'd say that there would never be more than 1000 - 2500 cycles per year of the valve on the seat.

A colleague asserts that this kind of cyclic discharge is not an appropriate use of the PSRV, and that we risk damaging the hard metal seat. I have said that I'll check the temperature of the vent piping, and if the valve starts to leak, I will simply service/replace it.

MY QUESTIONS are:
1) Has anyone experienced this kind of cyclic discharge?
2) Do I risk damaging any other components? I can't imagine fatiguing the spring.
3) Other comments or suggestions?

Thanks

[reference thread408-34442 ...]

 

[Ehzin]

Has there been any calculations of the required capacity for this scenario and do you have a rated capacity for the valve? Do you have any idea of the type of trim for the relief valve? Once the valve settles back down and reseats what does your steam system do? It could help to know if this is just coming off a very large header in comparison or this a small system.

I would guess it would be something related to an overcapacity issue. Depending on how "fast closing" your other valves are you could possibly have some form of pressure pulsations thru the line although I think that's supposed to be more relevant to rapid closing on liquid systems. Someone else will probably have better comments.

 

[LittleInch]

Without understanding the piping layout and the sizing of the valve it's not easy, but this sounds like your valve is significantly oversized for that condition and hence it is "chattering". I can only assume the PRV is sized for a different event, but the fact that it becomes isolated doesn't sound good to me.

Once or twice OK, but anything more than 2-3 times a year and this becomes an event your system needs to allow for.

Hence it would seem to me that you would want to set your main PRV maybe a little higher and install a valve properly sized for the flow, possibly using a pilot operated valve to get the accuracy up on the set point to say 310 or 320 psi.

I essentially agree with your colleague. Whilst you appear to dismiss potential failure this as an inconvenience if the seats leak, you're messing about with a vital safety device. Failures of such equipment can take on unforeseen consequences and maybe even fatigue the valve itself. 300 odd psi steam isn't something you want flying around a test lab.

You also want to reconsider the closure sequence if the line is becoming "over pressurised". Pressure safety Valves are there as a primary SAFETY device which if the system is correctly designed and operated should never go off. Normalising a safety event as part of what can be expected to occur is not a good way to go. Each operation of the valve should warrant an exercise in how can we stop this happening next time, not "It did its job, lets rely on it to work perfectly every time" way of thinking. This way of thinking was identified as part of the root cause behind the Challenger disaster when failure or part failure of one of the seals became part of what happened.... Just saying.

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

 

[btrueblood]

What LI said. Why not have the "fast closing valves" open a vent path, if this is the intent during a shut down. You say the PSRV is isolated when the valves close - is the relief valve isolated from the pressure generation source (boiler)? Is there some other PSRV on the source? If not, you are creating a bomb, and apparently kicking it around the room with each "unplanned shutdown".

 

[poetix99]

Thank you, respondents, particularly LittleInch, for your comments. The suggestion that I am "normalizing a safety event" got my attention, for which I thank you.

The safety valve is indeed sized for a different event which has not (yet) occurred. Although the trip events, with much less vented steam flow, were anticipated, the chattering behavior was not. I will consider the addition of a smaller relief valve with a lower pressure setting.

To clarify, and address questions/comments from all:
Within a length of approximately seventy feet of 6"-sch80 pipe & 600Class fittings (all good for >600psig), I have a Pressure Reducing Valve (PRV), and the subject Pressure Safety Relief Valve (PSRV) downstream of the PRV. The PRV reduces steam from a nominal pressure of 600 psig to something on the order of 100 psig (± 50).

The test equipment is rated for 300 psig for unavoidable reasons. The PRV is to normally assure the delivered pressure is less than the equipment rating, and the PSRV is to protect the equipment in the (extremely unlikely) event of the simultaneous failure of the upstream trip valve and the PRV.

Upon the occurrence of any of several possible events requiring an immediate shutdown of the test equipment, two "trip valves" (one at each end of the pipe run) close quickly, and basically isolate that 70ft length of pipe. The PRV will also close, albeit more slowly, and before it closes, the "settle-out" pressure of that total length of pipe will therefore be 325 psig (±25). All of the components within that length of pipe are rated for the maximum steam conditions from our boiler, and I am simply isolating it from the test equipment.
The boiler has its own safety equipment.

Again, my thanks.

 

[LittleInch]

poetix99,

Many thanks for your considered response which makes it all worthwhile.

It looks like the situation physically is better than described, but still needs some attention, IMHO. So basically it sounds like you're locking in two different pressures of steam and the Pressure regulating valve isn't closing fast enough? Maybe add a QA valve right in front of the PRV? You seem to have got the drift of my argument.

The analogy I thought of later ( I like analogies) is this is like anti lock brakes on your car. Great safety addition and have avoided many incidents, but if you find yourself feeling them come into action more than once or twice a year it tells you you're driving too fast or too close to the car in front....

When you figure it out, be sure to let us know, even if it's months later.

LI

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