PSV reliving rate as pump discharge valve fail open ?

[silebi66]

Dear Eng-tips professional :

I would like to find out the PSV-B relieving load as
(case1.) FV-1 fail in open position, FV-2/3 keep normal flow.
(case2.) FV-1/2 fail in open position, FV-3 keep normal flow..

Refer from pump system curve.:
DeltaPA is corresponds to normal operating conditions, and deltaPB is corresponds to relieving conditions. Since every control valve has its own Rated Cv that doesn't change (it's inherent in the valve itself),
In this case, the operating point @ relieving condition has to shift to the left to keep the same value of Cv @ normal condition because deltaPA @ normal condition is larger than deltaPB @ relieving condition.
I assume it will result in having a Qeff flow (see diagram). It seems not make sense that this result in less flow @ relieving condition than normal flow (Q 100%).

As the pump with 3 lines going out of it (equipped with FCVs 1, 2 and 3) and FV-1 suddenly fails to open, FV-2 and FV-3's % opening will adjust to maintain their flow.
Unfortunately, I don't quite understand how to solve this. I guess it must to be considered the hydraulics for all lines, and how the pump to be behaved (pump curve).

First, assume that when FV-1 fails, FV-2 and FV-3 will still keep its (normal) flow. But, I don't know how much flow increase in FV-1 when it fails open. Someone suggest that I should assume a flow in FV-1, then calculate the equivalent pressure drop across FV-1 (using FV-1's Rated Cv), then pick the new pump operating point (because we increased the total flow across the pump), then find out the required pressure upstream and downstream of control valve FV-1.

But, I am confusing about what is the system curve @ relieving condition moving upward or downward comparing the normal system curve,and the CV for FV-1, FV-2, FV-3 has to be the same value or not.

Please help me out in these messy wrong logical thinking here.

Thank you so much for your help.

 

[LittleInch]

I have to ask if this is a real problem or some sort of test as there is so little real data here and no tag nos.

Your system design doesn't make sense and for starters you shouldn't have a manual bypass valve around a control valve. If you do then sure as the day is long some lunatic will open it regardless of any locked closed or operating procedures. If there's no valve they can't open it.

Your pressure relief flow needs to be based on max pump flow with whatever friction losses occur at fcv fully open and presumably with outlets on vessel b blocked.

The lack of instrumentation , shutdown valves and simplistic sketch here is also leading me to suspect that this is not real life. What are your valves controlling on?

Your system curve remains the same regardless. Your system curve in relieving condition is incorrect and doesn't exist. Forget it.

What changes is the vertical line from the pump curve to the system curve caused by the pressure (head) loss from the control valves. As it reduces as you open up the control valve you move to the right, not the left.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[silebi66]

Dear LittleInch

Thank you very much for your comment, and I agree with you, but this example is just revised from Aiche continuing education course presentation.
Please bear with me, cause I still have trouble to understand this relieving capacity calculation issue.

The flow rate through FV-1 have four combination
(1) Q1=FV-1 @design opening + Vessel_B @normal operating pressure
(2) Q1=FV-1 @fail opening + PSV_B @ 5 barg
(3) Q1=FV-1 @design opening + PSV_B @relief pressure
(4) Q1=FV-1 @fail opening + PSV_B@relieving pressure
As the relieving pressure @ Vessel B must be greater than operating pressure @ Vessel B, FV_1 fail opening percentage must be greater than design opening percentage,
,so Q2 > Q1 > Q4 > Q3. Is any deduction above wrong or not ?

Please give me any comment and suggestion. I appreciate your kindly help.

 

[LittleInch]

You're doing this all wrong. Your flow items above mix flow and pressure all in the same equation

First decide what is your releiving case and conditions - Is it with all outlets (other than PSV) on vessel B blocked? or do they contiune to flow? What is your max inlet pressure and flow if the PSV on vessle A is set at 10 barg and the pump is still pumping??

Then work out whether you are looking at one, two or all three control valves going fully open.

Your max relieving rate comes from the pump, its upstream max pressure and its pressure differential. From the max pressure of the pump, considering that vessel B will be at 5 barg when relieving, you can then calcualte how much flow you can get through any combination of control valves which are fully open that you choose. Fully open, each valve could have a different flow on the assumtion that FV1,2 and 3 are all different. Thus the relieving flow rate will be a little less than the pump max flow if the pump inlet is max 10 barg due to the upstream PSV. How much less you can calcuate based on valves max open CV and the max differential pressure (10 barg + pump)-5 barg.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[silebi66]

Dear LittleInch

I really appreciate for your reply and suggestion.
Please forgive my stupidity.
I don't quite understand you said the max relief rate is based on calculation on the valves max open CV and the max differential pressure (10 barg + pump)-5 barg.

Could you please explain why the max differential pressure is 10 barg + "pump" - 5 barg ? And is the "pump" stand for the maximum pump discharge pressure or something else?

really appreciate for your kindness.

 

[LittleInch]

It is a common misconception that pumps deliver a certain pressure or head. They actually deliver a certain differential head. Thus if your pump is delivering say the equivalent of 3 bar differential, but the inlet pressure is 7 barg then the output will be 10 barg. However if the inlet pressure is only 1 barg then the outlet is only 4 barg.

Given that your inlet vessel has a relief valve set at 10 barg, then this is the highest inlet pressure the pump will see. Add on the pump differential head and you get the highest pump output pressure. Your downstream vessel has a relief valve set at 5 barg, therefore the differential pressure across the fully open valve is (10 + pump) - 5.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way