Sizing relief valves based on regulator failure modes 5

[lhundley]

If one is sizing a downstream relief valve based on pressure regulator failure conditions, can anyone tell me what flowrate/worst downstream pressure to assume for the failed regulator case? Haven't seen any real information on this. Am I to assume the worst a regulator can do is go full open to its mas downstream outlet pressure?

Lucas

 

[Bill3752]

Lucas, not sure of your specific case, but I usually handle a regulator the same as a control valve letdown. The required RV rate would be computed using the flow through the regulator under relief conditions. This means you are allowed to compute the rate based on the dP = upstream - OVP (=set point x 1.1). If you have a constant consumption of material, you could also net that out; ie. if the regulator fails, then some of the material would continue to be "consumed". A more conservative approach is to base the flow calcs on dP = upstream - set pressure, and/or no credit for consumption. Hope this helps.

 

[don1980]

I get the Cv of the regulator, and calculate the flow the same as you would for any other open valve.

 

[moltenmetal]

don1980 has it right in my opinion, unless you're dealing with a very special sort of regulator.

The regulator failures I've witnessed have been many and have happened as follows: the operator attempts to adjust the setpoint. The regulator internals "stick", such that outlet pressure no longer responds to changes in setpoint. In frustration, the operator whacks the regulator with the wrench, un-sticking the regulator. The regulator now has a setpoint greatly higher than intended, and initially at least it dumps gas or liquid into the system through its fully open orifice (i.e. it is wide open with flow determined only by Cv). A relief valve to protect against that event would seem to need to be able to keep up with the regulator stuck wide open, at least for a time.

 

[BenThayer]

i agree with the Cv of the regulator at wide open but i also look at the line size and dP in the piping.

it makes no sense to ignore that you might have 100 feet of 0.5" piping on the inlet to a regulator with a 0.25" orifice followed by 25 feet of downstream piping.

i have seen folks take the 0.25" orifice and treat it as if that is the only restriction in the system. If it was a 2" line feeding the 0.25" orifice and a 4" line leaving it, maybe but not relatively small piping where the dP in the system in a "fail open" case results in significant line loss......

 

[poli60]

I agree with BenThayer.
The flowrate calculated thru the fully open valve will be very high and so the pressure drop that it would be generated in the d/s system (ie piping between control valve outlet and PSV inlet) will be higher than the actual one.
So a trial-and-error calculation is needed to obtain the real flow.
This kind of problem can also be managed thru a dynamic simulation.

 

[asade]

Let us consider this scenario: A PV is installed on pump discharge line to regulate the discharge fluid pressure from 526 psig to 360 psig with a throughput of 70,000 bopd. If for any reason, the PV fails, it is expected that excessive pressure would be experienced in the downstream pipeline. To avoid this, a PSV would be installed.
What condition should be used to determine the relief flow rate to size the PSV?
Someone suggested that the relieving flowrate should be the total throughput from the valve. Is the condition right?

I am what I am by His grace

 

[TD2K]

Asade, you need to look at the hydraulics for the pump, the source pressure, the pump curve and then the losses in the piping between the pump and the PSV.

Basically, the PSV needs to pass enough fluid that the resulting head added by the pump minus the line losses do not result in more pressure than allowed by the applicable piping codes you are designing this system to.

 

[don1980]

Asade, please start a new thread with this question. It's a good question but it's completely unrelated to the topic of this thread.

 

[MortenA]

ISO 23251 (API 521) sec. 5.10.3 states that you should assume the worst case (unless you have several parallel feeds i which case you can assume that the remaineder og the valves remains in normal operation). So for an inlet valve uassume it opens fully.

SO your scenario could be: Outlet valve (or ESD) is closed (as a part of your operation). Incident: inlet valve fails to close. Here assume that the valve not even fails to close but opens fully. However, deduct the the pressure in the vessel is higher than during normal operation (as mentioned before ST+10%)

Best regards

Morten