PSV's Rated Flow vs Required Flow in Control Valve Failure Case 3

[kiade]

Please give me your idea:
In case of PSV’s sizing for "control valve failure" you have to calculate PSV’s required relieving rate based on maximum flow can be passed through the control valve which it is calculated from maximum cv of the control valve. As everybody knows, the orifice’s selected area is more than the calculated area and pressure drop calculation for inlet/outlet (tailpipe) should be done for the orifice’s rated flow. The question is “why do we have to use PSV’s rated flow for pressure drop calculation where, in fact, there is no much flow exist to pass through the control valve?” Remember we have already calculated the maximum flow that it can pass through control valve with maximum cv. Thank you for sharing your idea.

Kia

 

[Chance17]

When a safety valve activates, it goes fully open (there is a long side discussion why it goes fully open. just accept for now that it does).
If the inlet or outlet pressure drop is too high the valve will "chatter" -i.e. repeatedly slam shut and then open.
Chatter can damage the safety valve. One type of damage is the valve will not work when needed. A second type of damage is the valve does not shutoff.
For larger valves, Chatter may shake the entire structure and cause negative things to happen.
If you have ever been next to a large safety valve chattering, it will be one of life's most memorable experiences.
Short answer,the inlet/outlet pressure driop rules are to avoid chatter and safety valve damage.

 

[kiade]

Chance17, thank you for your answer but let me be more clear: I believe for PSV sizing in control valve failure case, considering the rated flow for PSV’s inlet/outlet pressure drop calculation, is not necessary because we have already considered it.

Kia

 

[OilToil]

I think Chance17 gave you the correct answer, and then went from there and gave you some extra things to consider.

Consider the system from the perspective of just the PSV - when the PSV inlet side reaches the set pressure, the PSV will go fully open, and the flow rate will be governed by the pressure at the PSV inlet, the orifice area, and the downstream pressure.

This rate can be higher than your feed rate. It just means that the PSV will flow fast and then shut. (I do wonder in scenarios like this - my gut feel is that you often don't have a long enough continuous event to create e a fully formed pressure profile in the downstream side, but that is what me must design for)

One complication in this is that you maybe have sized your PSV for your control valve failure based on an 10% overpressure allowance. This would have allowed a smaller orifice to achieve your flow, but at the same time, if you never get to that 10% overpressure you might be able to design the downstream side based on only the rated flow at set pressure.

What issue do you have with designing your downstream side for the (presumably) slight difference between rated capacity and actual flow? Selecting a different PSV type (eg. bellows) can mitigate back pressure issues.

 

[don1980]

Here are my comments on your post:

1) It isn't necessary to size the PSV based on the maximum sized control valve trim. I would use the trim that's installed in the valve. A P&ID note can be used to caution against changing the trim without checking the PSV size.

2) This might be good application for a modulating pilot operated PSV, if this system doesn't have much volume. Otherwise you might have a problem with rapid cycling, and/or chatter. The use of a modulating value allows you to calculate the pressure drops (inlet & outlet) based on the required flowrate rather than the rated flow.

 

[jamesbanda]

Don,

To followup.. have you seen guideance on when chattering maybe an issue..
e.g. % of > valve rated flow / required relief rate .. e.g. .1.5 is too much or 1.8 is too much.. 1.2 requires pilot valve..etc..

 

[don1980]

James - there are a number of factors that contribute to chatter. Of course, the one we're most familiar with is excessive inlet line losses. In vapor services, the factory blowdown setting is typically 7-10% of set pressure. If the inlet losses exceed the blowdown, then the valve will chatter. That's the basis for the 3% rule.

Your post mentions another cause of chatter - that is, oversizing the PSV. Unfortunately, there's no way to define how much oversizing is too much, because it depends on installation-specific factors such as the capacitance of the system, the accoustic length of the inlet piping, and probably other factors that are currently unknown. Also, understand that oversizing is inevitable, unless an error is made in the sizing calculations. All PSVs are going to cycle, and there's no objective way to distinguish non-destructive cycling from destructive chattering. For large vessels (high capacitance) this is probably not a great concern. If you have a relatively small vessel, and the consequences of a loss-of-containment are high, then you might have a cause for concern which justifies an alternate design (modulating pilot valve, or staged conventional valves). That's because the PSV cycling frequency is likely to be high enough for it to be classified as chatter rather than non-destructive cycling.

 

[kiade]

Sorry guys I was busy with other stuff during the past few days. Thank you Chance17 , OilToil, don1980 for your answers.
Then selected orifice size, upstream and downstream pressure are only three factors which you should consider for PSV’s inlet/outlet sizing regardless of how much flow is available to relief in order to avoid CHATTERING. And the same thing in case of liquid thermal relief where we don’t have enough flow to relief because the source of flow is only very small amount of excessive liquid due to thermal expansion. I believed the chattering in liquid thermal expansion is NOT going to be issued because of small amount of relief flow but based on this, there is no exception for that. Am I right?

 

[kiade]

don1980 / Cahnce17 /OilToil
Do you think the thermal expansion and control valve failure (as I mentioned above) will have the same issue (chatteing) regardless of the rated flow technically is NOT available?

 

[don1980]

Yes, ambient heating causes very low thermal expansion relief flowrate. Consequently, chatter is not a concern when the valve opens during such a scenario.

 

[kiade]

Thank you don1980.
So for the thermal expansion relief scenario, there is no concern about pressure drop across the PSV's inlet line (max 3% of set pressure) and you can always ignore it. (No need to calculate)

 

[don1980]

In most cases of thermal expansion due to ambient heating, the valve opens to release a small amount of liquid, the pressure immediately falls below the valve's blowdown pressure, and the valve recloses. In such cases the flowrate through the valve is typically far less than the valve's capacity, and the event is over with long before the valve can reach it's full lift.

The 3% rule is aimed at ensuring that the valve operates in a stable manner when it's relieving at it's rated conditions (rated capacity at 10% overpressure). For thermal expansion cases like the one we're discussing (expansion due to ambient heating) the valve never gets to 10% overpressure, nor to it's full capacity. Thus, there are no concerns about instability (chatter), and there's no need to calculate inlet losses.

This is the basis for paragraph 5.14.2 in API-521, which discusses the common practice of installing a 3/4"x1" PSV for common thermal expansion cases without the need for sizing calculations.

That said, you shouldn't go out and practice this without using good engineering judgment. Your question is phrased in absolute terms. This is a general practice that's applicable in most cases, but there are always exceptions to the rule. For example, if you have a 12" liquid transfer line that's 5000' long, then you ought to do a sizing calculation to determine the required relief flow rather than just installing a 3/4x1" PSV. And, if the required relief flow was more than say 40% of that valve's capacity, then I'd do the inlet calculation and apply the 3% rule.