11 replies to this topic

[Arora]

Hi All,
I am an Instrument Engineer for an EPC compnay. On one of my applications, I need to size a relief valve for a Lube Oil Tank ( Design Pressure = 25 psig ) which is being maintained at 20psig through a blanket air system.
I have used two criterion for sizing:
1. Failure of regulator on the blanket air line which causes the over pressurization of the vessel. In this case I only come up with a 1/2" x 1" relief valve.
2. When I size it for external assuming that the entire tank (1000 gallon) is full, I come up with a 4" x 6 " valve. This doesn't seem right as all the other existing tanks have a 1/2" x 1" relief valve. Am I doing something wrong by sizing it for external fire ? Also, if the vessel is double walled and we are using the equation q=21000FA^0.82, can we take some kind of credit on F (not assuming it's 1 as for uninsulated vessels). Do I have to consider air between the walls for thermal conductivuty purposes ?

I would really appreciate your replies. Please feel free to let me know if the question is not clear.

[Nirav]

hello,

In my opinion, we can take credit on F since air is considered to be a very good thermal insulator. Based on air conductivity, you can reduce value of F significantly by using guidelines provided in API.
I think the required rate should reduce to a value lower than value based on your first criteria and you would be able to consider 1/2" x 1" PSV.
(I'm guessing this because of similar existing tanks you are refering.)
Did you check the calculations for those similar existing tanks? It may give idea.

Any other opinion?

[Arora]

Nirav,

Thanks for your reply. After taking credit for F the flow rate dropped significantly and 1/2" x 1" seems to be a good choice. What confuses me is that API 521 does not cover this anywhere ? I mean how to size relief valves for double wall tanks ?

ANy comments ?

[djack77494]

I would respectivefully suggest that you include your vessel dimensions and orientation so that others can offer what might only be a "gut feeling" on whether or not your PSV size seems reasonable. It's very hard to develop any opinion on whether 1/2" x 1" or 4" x 6" is reasonable without this most basic information.
Doug

[Arora]

It's a 1000 gallon tank with the following dimensions:
Diameter : 4 ft
Length: 11.5 ft

Operating Pressure: 15-20 psig
Design Pressure: 25 psig
PSV set pressure = 25psig

[pleckner]

So far in my research, I haven't found much on double walled tanks except for the following from API 2000. This is for above ground refrigerated tanks. However, there is a significant similarity in the design between this type of tank and one for other applications as the the inside tank is usually at very high pressures. I'll keep looking but I've found nothing in ASME or in most of the other API publications that I have, which are quite a few.

"The heat input from a fire initially causes the vapors in the space between the walls of a double-wall refrigerated storage tank to expand, and the heat input also causes the vapors in the roof space of a double-wall tank with suspended-deck insulation to expand; however, it may be several hours before the increased heat input to the stored liquid causes a significantly increased vaporization rate. The venting requirements
for handling the increased vaporization may be small compared to the requirements for handling the initial volumetric expansion of the vapors.

Because emergency venting for a double-wall refrigerated storage tank is complex, no calculation method is presented here. A thorough analysis of the fire relief for a double-wall refrigerated storage tank, including a review of the structural integrity of unwetted portions of the outer wall, should be conducted."

Notice that even API recognizes that this is not a straight forward calculation.

I guess using API RP521 and taking into account a fairly large deration factor is pretty reasonable.

[Arora]

Mr Leckner
Thanks for posting the reply. I re-calculated the F factor in the API 521 equation keeping in mind the thermal conductivity of air ( at mean temperature ) between the double walls and the F factor comes out to be 0.021. This makes a big difference to what I had initially used F=1(for uninsulated vessels). The new relief valve size I come up with now is 1/2" x 1" which seems more reasonable.

Thanks.

[djack77494]

Thanks for your quick response Arora. I did a rough check on a fire case (only) for a drum with the dimensions provided. Basically it was just a

q = 21,000 * F * A^0.82

type of calculation. With F=1, I arrived at a heat input of about 1,500,000 btu/hr. With a latent heat as high as 1000 btu/#, which is higher than any system I've ever run across, so it will result in low vapor generation, I calculated 1500 #/hr of vapor generated. For the sample calc I had set up, this resulted in a "G" orifice, available with 1-1/2" or 2" inlets and 2-1/2" or 3" outlets. Your latent heat will probably be quite a bit lower, and you have to consider other cases such as regulator failure, etc., so your PSV may be larger.

As a side point, I cannot imagine a practical situation where I would be considering a double walled pressure vessel for lube oil. If your tank is insulated, I would still use a value of F=1 unless measures had been taken to ensure the insulation would remain intact despite firefighting efforts. I believe the applicable codes do a good job of discussing this. Having said that and read Phil's response regarding double wall vessels, I'm glad I've never had to do that type of calculation,
Doug

[Nirav]

Mr.Arora,

I have couple of more specific concenrs. You can check them in your project specific documentation. In one of the projects I worked, following requirements were considered (project specifications) for "fire case" calculations. I'm not sure if it is related to any of the standards or practices. I'm looking for others' opinion on this.

[1] In no case, F should be considered lower than 0.075.
[2] For fire case, safety valve should not be less than 1" x 2", D orifice.

I'm not suggesting you to follow these. Just trying to put a point of concern.

Thank you,

[pleckner]

I can't find these restrictions mentioned anywhere in the standards or codes I have.

The lowest F value stated by NFPA 30 is 0.15.

Perhaps the PSV size restriction is per your client. However, you can't get much smaller and still be able to relieve anything significant in the vapor phase. As far as I'm concerned, the PSV size will be whatever it will be and practically speaking, I can't believe it will calculate out to anything smaller than a 1 x 2!

[Arora]

Thank you all for the replies. I had been offshore for last week and never got a chance to read any of these replies.
Thanks again

[proinwv]

Hi,

I am just curious, but why are you blanketing at 20 psig?