Scaling relief loads by orifice size?

[Amphoteric]

Hello,

I've just started working through my first relief project and am getting conflicting advice sizing relief valve lines. Allow me to explain with a hypothetical example, because I'd like to learn the general principle/approach instead of just getting through a single problem:

Say I have several relief scenarios. I perform my calculations with the help of API 520 and determine the following loads and minimum orifice areas:

w1 = 100 lb/hr, a1_req = 0.2 in2
w2 = 1000 lb/hr, a2_req = 4 in2
w3 = 1200 lb/hr, a3_req = 3.7 in2
Conclude: Scenario 2 governs.

Then, I find a table of orifice sizes like

http://www.contractorsunlimited.co.uk/toolbox/reliefvalveorifice-curtiss-wright.shtml.

Since I need 4 in2, I take the next larger size, an "N" orifice (a_act = 4.340 in2).

My next step is to size the relief valve's inlet line (from vessel to PSV), it's outlet lines (from PSV to flare header), and to report my loads to the person designing the inter-unit headers and the flare burner. If I'm using a conventional PSV, I'll want to size the lines so that during the worst possible scenario I keep the inlet line pressure drop below 3% of PSV set-pressure "Pset", and limit the backpressure on the relief valve to 10%Pset.

My question is, what load do I take when sizing the inlet and outlet lines? And what load do I report to the guy designing the flare system? Are they different?

I can see at least three options:
1) wreq = w2 = 1000 lb/hr. The load that provided the largest orifice size.

2) wreq = w2 * a_act/a2_req = 1085 lb/hr. The governing scenario, increased because the actual orifice is larger than the minimum size I calculated using API.

3) wreq = w1 * a_act/a1_req = 2017 lb/hr. The orifice is so much larger than what is required that scenario #1 actually governs.


Which is it? And it gets more complicated if the scenarios are of different phases or wildly different fluid properties. Or if scenario #1 was the fire scenario, and so generation perhaps can NEVER be as high as 2017 lb/hr...

Do I even need to considering having two relief valves, one small valve that pops early to relieve scenario 1, and a larger valve that pops later in case of scenarios 2 & 3?


Sorry for the lengthy example, but I wanted to raise all of the issues that I have been considering at once.

 

[CHD01]

FIrst , you need to review in detail the ASME CODE Section VIII Division 1, part UG for relief valves. You also need to review API520 and API 521 before you are ready to size relief valves.

Now I will answer some of your questions which will only generate for you more questions as you realize even more how much there is yet to learn.

The worst case is always the largest relief area required. However, the largest required relief rate calculated will not alwayss result in the largest relief area required. This is because the fluid relieved and relief conditions are not always the same for all relief cases; and some may be liquid or vapor or 2-phase flow.

Next, the flow that is used to calculate the line size for both the inlet and the outlet of the relief device is the relief valve rated capacity for the fluid being relieved. There are only a few exceptions to this, one could be liquid thermal expansion where the relief valve capacity is impossible to achieve.

Also, the flare load (required relief rate) is the combined rate that could occur simultanelously from all relief devices entering the flare header - note that in this case we used the required rates, not the relief valve capacities.
For example, for a fire you may have 20 relief valve going to the flareheader but the fire case will only involve maybe 5 of theose 20 devices.

One more question you had, sometimes you do want yo install more than 1 relief valve for a singe vessel. It is your judgement based on many considerations. Do you have both a liquid and a vapor and do they need seperate valves? Maybe you have a column with packing and there are no liquids to relieve but you are concerned about the packing plugging. So you decide to install 2 valves - 1 ahead of the packing and one after the packing atop the vessel as the Code requires.

You are on the right track by asking te right questions, just keep working at it.



The more you learn, the less you are certain of.

 

[Amphoteric]

CHD01, thank you for the excellent reply. I believe I understand what you are saying. Even though w3 > w2, scenario 2 controls because a2_req > a3_req. The lines will be sized using wreq = w2 * a_act/a2_req = 1085 lb/hr.

The flare header will be sized assuming w1, w2, or w3, is relieved, depending on the scenario. (So if w2 is a power-failure scenario, it will be w2 = 1000 lb/hr extra to add to the other power-failure-loads).

It seems strange to me that you don't scale up the loads going to the flare header. Isn't the valve actually going to pass 1085 lb/hr at some point?

 

[CHD01]

While scaling (ratioing) the areas works much of the time to get a valves maximum flow, remember that you really want the relief device capcacity. Scaling will not always give you that. This is because the relief valve is always certified (rated) at 10% overpressure for section VIII valves (3% for section I); amd if the worst case is a fire case this orifice is sized at 21% overpressure. Therfore if you scaled an orifice sized for 21%, your scaled up flow would be in error.

Capacity of the valve is used because the inlet drop must be based on this to determine if the 3% rule is met to avoid mechanical chatter. For a relief vent header where flows are combined you are not checking for chatter you are onluy interested in the actual pressure drop for the actual combined flow.

I always try to get the valve capcacity freom the vendors capacity tables. If not available I then reverse the relief valve area formula to calculate the capacity for the correct % overpressure.

The more you learn, the less you are certain of.

 

[JoeWong88]

Amphoteric,
First you have define / guestimate what type of PSV you will use. Check API. There is section discuss about relieving flow or rated flow to be considered.
Read more in "Consider Rated flow or Required Relieving Flow ? ".

Strictly speaking you have to ensure all condition is within the 3% requirement. In certain event, you may exceed the 3% requirement if the PSV is having high blowdown.

JoeWong
Chemical & Process Technology