7 replies to this topic

[backlit]

Can anyone please explain the formula used in API 521 for the confined liquid volume subjected to temperature increase and consequent pressure increase. I am interesetd to know what the numerator term with the difference between cubic and linear expansion coeffecient. The denominator term for the steel wall subjected to the pressure increase.

 

I want apply this for case of a ball valve body cavity relief.

 

Sajit

[backlit]

Hello Sidd,

 

This link will show the API 521 formula and the formulas that I had entered. the heat transfer and the Cp does not figure in the formula. Does it get cancelled out?

 

https://docs.google....dlE&usp=sharing

 

I could not find how I should reply you. I therefore copy pasted your post below.

 

Sajit

 

"Simple explanation hope this wud clarify:

for liquid expansion case, safety valve will reliev the difference between initial & final volume which has expanded due to change in temperature. change in volume can be expressed as

The change in the units volume when temperature change can be expressed as


dV = V0 β (t1 - t0)
where
dV = V1 - V0 = change in volume (m3)
β = volumetric temperature expansion coefficient (m3/m3 oC)
t1 = final temperature (oC)
t0 = initial temperature (oC)

now API has just modified this formula, instead of term (t1 - t0), it has been replaced by heat transfer rate:

heat transfer rate Φ= mCp(t1 - t0)= V0 x density x Cp x (t1 - t0)

Replace this is original formala and you get API formula.

Hope this clarifies it."
 

[Art Montemayor]

What I have done in the past for this scenario involving the liquid-full cavity of a ball valve in the shut-off position is that I simply specify a drilled hole in the ball valve when I buy it.  The hole is usually 1/8" diameter and is drilled on the ball face that is on the upstream side.  The downstream side of the ball is the one that is doing the sealing.

 

By doing this, I keep things simple and mitigate the need of a PSV on the side of the ball valve (which presents a lot of problems and usually winds up damaged or broken off with time).

[backlit]

Art,

 

It used to be so, until they came up with DPE (Double piston effect) seats. A hole in the ball will make the sealing that a DPE seat will provide ineffective.

 

Sajit

[Art Montemayor]

Yes, that might be so if someone introduced "DPE" seats.  However, not ALL ball valves have DPE seats - at least not in the USA.  I know for a fact that many ball valve manufacturers fabricate ball valves with the relief hole already drilled in the ball.   All you have to do if you require such a feature, is to specify it.

[backlit]

I am afraid this one got specified as DPE as a carry over from FEED engineering. For both the seats. The body is already machined to receive the seats. To throw away the 18" body will cost and delay the delivery. if not could have changed one of them to SPE. Infact it is specifically negated in the specification for having a hole in the ball.

[Art Montemayor]

Then why not just apply a thermal relief valve.  No calculations are involved nor needed.  The amount of liquid volume expansion (due to thermal effect) is miniscule.

[backlit]

The cavity volume for a NPS 18 class 300 trunnion mounted FB ball valve is estimated at 30 L. The expansion is not miniscule if a temperature range of the 60 to 70 is taken. You may see the link above. Will some one be able to derive the API 521 equation from first principles to know whether it is correct to use for a valve cavity pressure increase.

 

Yes, I agree on the use of the relief or the hold in the ball. The contractor / manufacturer does not want to incur the cost associated and also there may be a schedule delay. There is also the concern that the relieving device cannot let off to the atmosphere. It has to be piped to the other side of the seat. This will have another vulnerable element, to external damage.