API 2000 Fire Case for FRP Tank 4

[KernOily]

Hi guys. Would you consider the API 2000 fire case to be applicable to a fiberglass tank for emergency venting? Or would you consider, since the tank will itself melt/catch fire when exposed to a fire, that the fire case would not be applicable to this type of tank?

Thanks guys! Pete

 

[don1980]

For FRP vessels exposed to fire, relief devices can't provide any meaningful protection. Thus, sizing for fire exposure is a waste of time and money. Worse, it actively misleads people into thinking that the relief device is actually providing fire protection.

When choosing to use an FRP vessel, and choosing the vessel's location, one should consider the fire exposure risks. If it's intolerable for this vessel to fail during a fire, then one must choose a different design, or a different location for installing the vessel.

Since you're a petroleum guy, I'm guessing this is an API 12P tank. If you look in the API 12P document, you find cautionary guidance regarding the use of those tanks in a fire zone, for the reasons I've stated above.

 

[KernOily]

Don - Thanks for your great reply. I assume your reasoning is coming from the idea that the heat input form a fire adjacent to the tank shell will damage/destroy the shell before the heat is transferred to the contents thus creating a large vapor relief scenario? Just trying to make sure I understand your point.

These tanks are ASME RTP-1. Thanks!

 

[don1980]

Fire sizing calculations are based on fire heat conducting through a metal wall into a liquid, which then boils. FRP behaves in a fundamentally different way than metal. The obvious difference is that FRP melts and burns. Even if we ignore that reality, the heat input equations in API 521 and API 2000 don't apply to FRP vessels. Those equations are based on metal vessel walls which will conduct high amounts of heat into the vessel contents (liquid). When the liquid on the inner wall starts to boil, the Hvap helps prevent the wall temperature from steadily rising. That's the mechanism by which relief devices protect from fire exposure. The heat input equations are based on empirical test data involving metal tanks. Thus, these heat input equations are invalid for vessels made of non-metallic materials, and that's especially true for a material like FRP which is more of a heat insulator than a heat conductor. For an FRP vessel, very little of the fire's heat gets transferred through the wall and into the liquid. Since that heat can't easily pass through the FRP wall, the result is that the wall temperature rises very fast - essentially none of this heat is passing through the wall into the liquid.

The bottom line is that applying the fire sizing equations to a relief device on an FRP tank is a gross misapplication of those equations.

 

[KernOily]

Don - Excellent and thank you. I thought the same as well so thanks for the confirmation.