Mitigation Options for Existing Oversized Relief Valve Installation

[Sirius P.Eng.]

Adequacy check of existing relief devices (2No. 8T10s set @17.7 barg; conventional)for LPG storage spheres confirms they are grossly oversized in the event of governing external fire scenario.

Using API model I calculated relief load of 79,833 kg/hr. Relief fluid z factor is around 0.6691 indicating non-ideal behaviour. However, I used API 521 sizing equations for ideal gas and selected a 1No 4P6 as adequate for overpressure protection.Certified rated capacity is 26.90%. Aspen Flare System confirms back-pressure is over 30% so a balanced type valve was selected. I reckon a dynamic simulation may even result in lower relief load. I am guided by the general acceptable overcapacity is between 25-30%.

I passed my results to a Contractor/Consultant for validation/verification.
Using the superior HEM Method, they calculated relief load is ~89,087.3 kg/hr. Now they proposed 2 options:


1. 1No. 6Q8 with valve lift restriction to about ~44.2% overcapacity.[restrict orifice area up to ~52 cm2 from full lift orifice area of Q (71.29 cm2)].

2. Insulate Spheres
Provide 1” thick foam glass (thermal conductivity is very less ~0.055 W/m∙K) conservatively as a fire proof insulation on listed sphere with environmental factor of 0.3 per Table 5 API 521 to mitigate oversize concern.

o Required Orifice area: (~26,730 kg/hr; 11.96 cm2)

o Selected Orifice area: (~ 41,160 kg/hr ; 18.406 cm2 considering full lift of 4L6 orifice; API area)

o Relief valve is ~54% oversized

In this case also, restricted lift will be required for lower 4L6 size.

Given this, which option should be considered considering a whole bevy of technical safety factors and cost.

Thanks.

 

[Latexman]

Will the insulation reduce product loses, such that you get an economic payback? If so, that needs to be considered too.

Good Luck,
Latexman

 

[georgeverghese]

This is a firecase relief only RV - just leave things the way they are. What if the pool fire were to be only a fraction of the API heat rate?
How do your calcs differ from the original firecase load calcs? Is it to do with drainage around the spheres?
If you use the firecase RV sizing routine in Pro/ II - Simsci, the firecase RV load may be even smaller.

 

[shvet]

P.Eng.]I am guided by the general acceptable overcapacity is between 25-30%.

Can you share proofs/links?

Given this, which option should be considered considering a whole bevy of technical safety factors and cost.

It is not a common practice to avoid an oversized relief device. What is a problem you are trying to avoid? Chattering? Cycling? Backpressure? Relief load to vent/flare?
Many of relief devices are sized for multiple case. Most of those cases have low relief load and therefore are oversized. As per my experience most of RD handling multiple relief cases are oversized as those was sized for one biggest case and was not multiplied&staggered.
Why oversizing of most RD is not a problem but this particular RD is a problem to be avoided by so expensive means?

For info:
1/ Note that para. 7.1.1 API 521 has no instruction for avoiding cycling. The same is valid for para. 2.4.2.2.4 CCPS "Guidelines for Pressure Relief and Effluent Handling Systems". Only a few companies I have met prohibited oversizing (ExxonMobil and BP) but threshold was 75% not 30% (see para. 7.5.1 DP XV-C or 10.1.a.1 GP 44-70 respectively).
2/ In your case additional small staggered RD may be an option.
3/ In your case stem lift stopper may become the cheapest option. Contact a manufacturer for plot "capacity vs lift". It looks like the same as transport pin on top of bonnet. Sorry I do not remember the correct name of this device.[/sup]