Design Pressure due to Deflagration

[MSL93]

It is common in my industry to design a flare stack for a pressure of roughly 7 barg that is solely intended to account for the pressure in the system if a deflagration event were to occur. In fact, several customers have this language put into their specifications. From my research, this looks to be a safe and simple way to handle the design of the system for deflagration. However, it is my understanding that the nature of a deflagration event is that it results in a rapid increase in pressure from the combustion of the internal gases that then propagates to the exit of the stack and into the atmosphere. If this understanding is correct, then this would be a rather transient load case where designing the stack for a static (sustained) design pressure of 7 barg may be unnecessarily conservative.

Do note that the stack system I am describing is downstream of all block valves and open to the atmosphere.

In general, this isn't a problem (designing the stack for an internal design pressure of 7 barg) for small diameter stacks where this pressure is relatively minor, but for stacks where some segments are over 10 ft in diameter, this 7 barg design pressure can result in a large increase in wall thickness and overall system weight. When all other load cases have operating pressures less than 1 barg, I want to be positive that I am not unnecessarily designing the stack for a sustained internal design pressure of 7 barg to account for deflagration.

Does anyone on here have an insights or references to add that may provide guidance to my scenario?

 

[LittleInch]

I've been looking recently at Hydrogen vent lines and some design those for 40 Bar DP.

Length of the flare line will have an impact as well as contents.

So 7 bar % 100 psi is almost certainly not a calculate pressure but just dimensions which has been shown to work.

A vent line which ruptures is worse than no vent line as it would spew harald vapour into confined areas and probably kill people. So a fair bit of conservatism is warranted.

 

[seuenergy]

Normally, the rate of pressure increase rise of hydrocarbon is roughly 7~8 times of deflagration, and 12~13 times for H2. So if the operating pressure of flare system is 0.5bar, so 0.5* 8 = 4 bar. Then 7 bar design pressure is reasonable.
There is also code from API, it has recommended valve. If my memory is right then the value is 5bar. Please correct me if I am wrong.

 

[georgeverghese]

You may wish to query the rational for these unusually large dia lines. Why is the max operating pressure only 1barg? Can you not allow for a higher operating pressure that will reduce line dia ? Or is there some limitation due to low design pressure of equipment connected to this flare system. Pilot operated safety valves will enable much higher operating pressures in the flare headers( up to 70% of set pressure if I remember correctly), and balanced bellows PSVs' up to 50% of set pressure, while conventional spring loaded PSVs' cannot handle flare header pressures more than 10% of set pressure.
A min of 100psig design pressure for flare headers in hydrocarbon service is widely used in practice, for cases where the flare headers may infrequently have to handle temporary accidental loss of purge gas.

 

[LittleInch]

Normally, the rate of pressure increase rise of hydrocarbon is roughly 7~8 times of deflagration, and 12~13 times for H2. So if the operating pressure of flare system is 0.5bar, so 0.5* 8 = 4 bar. Then 7 bar design pressure is reasonable.
There is also code from API, it has recommended valve. If my memory is right then the value is 5bar. Please correct me if I am wrong.

That's very interesting. Do you know which code from API?

 

[seuenergy]

That's very interesting. Do you know which code from API?

hi, sorry for my poor memory. API 521 suggests the minimum design pressure of knockout drum should be 0.345MPa. My thinking is that the flare stack should be not less than this value.