API 3% guideline - advice

[KenA]

Apologies if this has been posted sometime in the past, but can anyone help me with interpreting the API 3% guideline for inlet lines to relief devices?
API RP-520 Part 2 recommends that the pressure drop in the inlet line to a relief device should not exceed 3% of the guage set pressure of the device. It goes on to say that if the PSV is installed on a process line (as opposed to directly on the vessel), the 3% should be applied to the sum of the loss in the inlet line and the incremental pressure loss in the process line caused by the flow through the PSV. I have always read this incremental pressure loss as pressure drop @ rated PSV flow rate minus pressure drop @ normal flow rate. My question is, what should you use for "normal" flow rate?
The application I'm looking at is in a utility system, therefore the "normal" flow rate is variable. There is a nominal "average" flow and there is a max flow. The max flow is much higher than the average and the rated PSV flow is obviously larger than the max flow. Therefore the incremental pressure drop increase between average and rated PSV is very large. This leads me to question, why use the average? Can I use the max? But even this seems fairly arbitrary because the PSV could operate at any point in time.
Any advice?? thanks

 

[Montemayor]

KenA:

When you design the PSV to function according to the 3% API ruling under all operating conditions, you've done your job according to the intent and the spirit of the API's basis of design. This would mean, then, that you design for the maximum flow and that the 3% loss will never be exceeded.

The intent is to have a safe and protected situation under all operating conditions for the vessel or equipment in question. Otherwise, you only have a safe situation some of the time - not all the time. This is analogous to the "pregnancy" adage: you can't be a "little" pregnant; you either are or you're not.

 

[KenA]

Art,

Many thanks for your reply - your views are always valued.

However, I am still struggling. If the pressure drop in the process piping is an incremental pressure drop due to increase in flow due to PSV, you must have to haev a starting point. The flow has incresed from "x" to the rated flow rate of the PSV and as a result the pressure drop in the pipe has increased. Obviously the lower the number used for "x", the higher the resulting pressure drop increase. For a utility system, the flow rate could be practically anything from zero to the max. At low flow rates it becomes impossible to meet the 3% guideline. I just wondered whether anyone had come across this before.

 

[TD2K]

I've taken credit for the 'normal' process flow in sizing a relief valve though that's not my preference. I've never taken credit for the 'normal' dP through the piping if I understand you correctly.

In a case like yours (if I'm understanding it right), I would have the pressure drop for the PSV capacity plus the normal process flow (if the two are flowing through a common section of line) plus the pressure drop through the inlet leg of the PSV at its capacity. The sum of those two pressure drops have to be less than 3%.

In cases were we've exceeded that rule for existing systems, the client has sometimes decided to stay with the final pressure drop rather than replace piping or looking at other solutions. For new designs, we would upsize the piping or sometimes look at remote sensing lines if you have a pilot operated valve. We would also ensure the actual pressure drop didn't reduce the valve's capacity to less than the required capacity.

Hope this is clear.

 

[Guidoo]

KenA,

First I want to ask you to be a little bit more specific about your problem, e.g. what utility system are you talking about and where is the PSV located? What are the relevant pressures? etc. This is just to prevent getting a vague discussion.

Second, the important thing to bear in mind is that the 3% rule is there to prevent rapid opening and closing of the PSV. Pressure at the PSV inlet should not drop below 97% of the PSV set pressure just because the PSV has opened. If the pressure drop would be too high, you would get a rapid cycle of PSV opens --> Pressure at PSV inlet decreases --> PSV closes --> Pressure at PSV inlet increases --> PSV opens
When you know this, your problem can be solved without looking at the exact wordings in API RP 520.

 

[KenA]

Guidoo,
Thanks for your reply. I'll give you some specifics but the question applies to many instances that I can think of.
Fuel gas (operating 20 barg / 290 psig) is let down across control valve to 7 barg / 100 psig. There is PSV on downstream piping set at 10 barg / 145 psig. The control valve has a large turndown because "normal" flow is say 5000 kg/h (11000 lb/h) but maximum flow rate is much larger (say 20000 kg/h / 44000 lb/h) due to start-up or intermittent demands. Hence PSV has rated capacity of say 25000 kg/h (55000 lb/h). So my question is, for the common section og line between control valve and PSV, API asks for the incremental pressure drop. I'm reading that as the pressure drop for 25000 kg/h minus the pressure drop for normal flow. Am struggling with normal because that would imply 5000 kg/h, hence the incremental doesn't seem to buy me much. I may as well just assume the 25000 kg/h for the whole length.
Incidentally I diagree with TD2K's suggestion of adding the rated PSV flow to the normal flow for the common section because the flow can not be more than the PSV rated flow (in the case above).

 

[Guidoo]

KenA,

For your example, my advice is to locate the PSV as close as practicable downstream the control valve. The relief scenario would be fail open of the control valve.
The pressure drop over the PSV lead line (based on the actual flow through the PSV at relief conditions) should not exceed 3% of the gauge set pressure of the PSV.

I would say you should use a modulating type relief valve, to make sure that the flow through the relief valve matches the flow through the control valve minus the flow that is consumed by the fuel gas users.

Note that this is example is different from what is described in API RP 520 part II, that shows a PSV that is protecting an upstream vessel. Here we have a PSV that is protecting a downstream pipe network.

 

[roker]

hello,

since you want to protect a pipe system you can look on that as the vessel and the pressure drop is calculated for the line from this pipe system till the PSV at RATED flow of the PSV, with 10 barg set pressure you can usually meet the 3% rule.
by the way to what relief case the PSV is designed? because piping system can withstand 20 barg the fuel gas source pressure or there is a downstream equipment with a lower design pressure?

regards,
roker