Design temperature for flare system

[ixchawla]

I need to specify the maximum design temperature for flare system. The relieving cases includes an external fire case. During this scenario, the temperature inside the vessel (under fire) increases from 50 oC to start with and goes upto 350 0C,when final heaviers are left in the vessel. I would like to know the following:

1. Does the flare system to be designed for the fire case?I have seen varying practices being used in industry- a. Design based on normal operating conditions (plus margins, say 20-30 oC), b. Design for fire case.

2. If it is desired to design for the fire case, the design temperature needs to be established. Looking at the differential vaporisation inside the vessel, more than 90% (by moles) of the liquid gets vaporised below 130 oC, the balance 10% gets vaporised at much slower rate (~ 1/10 th by moles) with increase in temperature from 130 to 350 oC. Since the vaporisation is slow in the later part, the PSV essentially will start chattering (pop-action)thus giving some time for cooling the flare system with ambient temperature. Hence, I would consider that it is not necessary to design the flare system above 130 oC (after considering some margin, say 20-30 oC). I think designing based on 350 oC would be too conservative, and hence not required. To make it simple, the temperature drop across the PSV is not mentioned, however, it will be considered.

I would appreciate your views on this and correct me if I am wrong. Also, please specify if other ways of establishing the design temperatures are being used.

Thanks for your responses.

 

[aikmeam]

I find your interpretation on point 2 rather acceptable.

But also check the following:
a. Whether it's 150degC or 350degC, the piping would very well be below 150# rating, if it's a LP flare.
b. Maybe at the higher design temperature, your vessel thickness would still be lower than the min. required for mechanical.

And since the KO/seal drum would be located close to the flare tip (which is also hot), I'd rather worry about the heat conducted from the flame itself rather than the relief from PSV.

My point is, check the CAPEX impact on the differences. My guts tell me there's none. Why don't go higher then? Hope this piece of info helps.

 

[pmureiko]

Be sure and choose the correct design temperature (DT). The cost from a higher DT will be due to expansion loops, not the pipe material or schedule. The relief piping is probably large diameter and very long. In this case the problems from thermal expansion get worse with higher DT. You should to do a stress analysis of the piping which is a fixed cost. You may not need many expansion loops because of flexibility in your system from elbows and direction changes.

 

[roker]

Dear ixchawia,
since we are facing the same proplem: how did you design the flare line and how did you consider the temperature drop across the PSV.

REGARDS

 

[ixchawla]

The maximum design temperature for the flare system was estimated from the following criteria:

1. Normal operating temperature - Max 45 degC
2. Fire case (ESD system activated): Blowdown will limit the temp to within 100 degC.
3. Fire case (ESD system not activated): The test separator temperature can go upto 350 degC, however, the flare system fluid temperature will be within 250 degC. The metal temperature will be further lower due to heat loss to atmosphere.
4. The time required to heat the fluid (during fire) in test separator to 150 degC is more than 15 min.
5. The fire is estimated to last for maximum 15 min as per QRA study.

Hence, a design temperature of 150 degC for flare system was considered sufficient.

The temperature drop across PSV was calculated based on adiabatic flash.

Regards,

 

[25362]

To ixchawla.

Questions:

a. It appears case 3 means a "double jeopardy" approach.
Could a redundant ESD option be applied ?
b. Is there a chance for an exothermic chemical reaction of the contents ?

Thanks for your comments.

 

[ixchawla]

No reaction is expected to occur for my case. Its a good idea to analyse this on case by case basis.

When equipment/piping safety is considered, generally we donot rely on instrument system unless they are designed based on the availability requirement(eg HIPPS). Hence, I donot consider this as a double geopardy case. Let me know if someone has a different opinion on this.

 

[jcaiken]

Check your flaring scenarios to make certain they are applicable.

Flare header flow due to fire case typically requires the vessel to be isolated to generate the relief pressure, e.g. following an ESD or maintenance. A fire without isolation may not cause the RV to operate.

Flow through the RV to flare will have upstream temperatures at the boiling temperature of the liquid in the vessel and downstream temps dependant on the JT cooling across the RV. Emergency depressurisation of the vessel because its temperatures are rising too high (all the liquid has vapourised) will marginally lower the boiling point of the remaining liquid but significantly reduce the dP (and associated JT cooling) between the vessel and flare header. However, the mass flows will also be decreasing and the heat transfer in the piping will become a more significant factor in determining the piping temperature.

The transient nature of this type of problem means that it may require use of simulations such as the dynamic depressurisation model in HYSYS to get a 'better' answer.

 

[CHD01]

What is your pph flowrate?

The more you learn, the less you are certain of.

 

[CHD01]

Sorry, I also need pipe diameters and lengths; plus specific heat of flowstream if you have it.

The more you learn, the less you are certain of.