5 replies to this topic

[Mahdi Malek Alaie]

Dear All

 

We are supposed to design a " Flare Gas Recovery and Utilization" unit for a gas refinery. The capacity of FGRU is 11.5 MMSCFD. 

The existing flare system consists of three lines: HP, MP and LP.

Could we use a " Liquid Seal Drum" for sealing the HP path. the total back pressure of the HP flare header is 4 barg.

Some of standards such as API 521 recommend the seal drum while Total: GS EP SAP 262 didn't say anything about seal drum and recommended the " FOV" (Fast opening Valve) instead of the seal drum.

Furthermore, is there any design procedure for seal drum except API 521?  

[SawsanAli311]

The use of liquid seal drum is acceptable by various design guidelines as it is a mean for prevention of air ingress, flash back and maintaining a positive pressure such that all vapor are routed towards the flare recovery system. However, our current experience with liquid seal suffers from corrosion problems as well as possibility of blockage of flare header in case of liquid carry over. The use of liquid seal allows the design pressure of the flare system upstream to be 3.5 barg while the liquid seal and downstream the stack shall be minimum 7 barg in order to protect from the effects of internal deflagration and possible detonation. 

 

Accordingly, as an alternative sealing mechanism, flare system of 7 barg with continuous purging can be used in conjunction with a quick opening valve and a bypass buckling pin RV set at a higher pressure point than the QOV to cater for the case of failure of QOV to open. This has been implemented in new projects in order to avoid the challenges associated with liquid seal drums. 

Edited by SawsanAli311, 12 September 2018 - 01:05 PM.

[demank]

... The use of liquid seal allows the design pressure of the flare system upstream to be 3.5 barg while the liquid seal and downstream the stack shall be minimum 7 barg in order to protect from the effects of internal deflagration and possible detonation. 

 

Accordingly, as an alternative sealing mechanism, flare system of 7 barg with continuous purging can be used in conjunction with a quick opening valve and a bypass buckling pin RV set at a higher pressure point than the QOV to cater for the case of failure of QOV to open. This has been implemented in new projects in order to avoid the challenges associated with liquid seal drums. 

 

Dear SawSanAli,

Would you like to share with us the reference of minimum 7 barg for the stack in order to protect deflagaration/detonation?

Thanks a lot.. 

[SawsanAli311]

In my company we follow Shell design guideline which has recently been aligned to the API 521 code in 80.45.10.10- Amendment supplement to API-521 Design of Pressure, relief, flare and vent system. The distinguishing of the 7 barg is clearly explained in conjunction with the absence of liquid seal system. 

 

In general terms,  it is generally known that the pressure generated by deflagration is around 8-10 times the absolute normal operating pressure of the system. For detonation, it is around one order of magnitude over the one for deflagration (unstable detonation developed in the transition zone between deflagration and detonation). If you see API 14 J, the dictated pressure for flashback protection is 125 Psi which is around 8.6 barg. This is another conservative value to be used for setting the flare system's design pressure when a seal is not there.

Some companies tend to set a minimum of 15 barg for the design pressure of the system to withstand the pressure waves generated from deflagration. 

Edited by SawsanAli311, 13 September 2018 - 04:24 PM.

[Mahdi Malek Alaie]

Dear SawsanAli

Thanks for your reply. So as i understand , the implementation of the liquid seal drum is not related to HP or LP (different Mach.No). it is sufficient to have a proper design. am i right?

In my company we follow Shell design guideline which has recently been aligned to the API 521 code in 80.45.10.10- Amendment supplement to API-521 Design of Pressure, relief, flare and vent system. The distinguishing of the 7 barg is clearly explained in conjunction with the absence of liquid seal system. 

 

In general terms,  it is generally known that the pressure generated by deflagration is around 8-10 times the absolute normal operating pressure of the system. For detonation, it is around one order of magnitude over the one for deflagration (unstable detonation developed in the transition zone between deflagration and detonation). If you see API 14 J, the dictated pressure for flashback protection is 125 Psi which is around 8.6 barg. This is another conservative value to be used for setting the flare system's design pressure when a seal is not there.

Some companies tend to set a minimum of 15 barg for the design pressure of the system to withstand the pressure waves generated from deflagration. 

[SawsanAli311]

Hi Mr.Mahdi, 

the main function of a sealing system is not related to header's being segregated by high or low pressure. It is primarily to prevent air ingress as well as flashback inside the flare system and maintain a positive pressure such that all vapors are routed towards a vapor recovery system (if applicable). 

In case no seal system in place, the purge rates requirements will have to be carefully increased in addition to maintaining the design pressure of the complete flare system to a minimum value of 7 barg such that it can withstand any impacts generated from internal deflagration and possible detonation (unstable transition from deflagration). 

Edited by SawsanAli311, 21 September 2018 - 10:20 AM.