11 replies to this topic

[kc2010]

Dear Experts,

 

According to API-14C we have to provide two independent protection layers for each identified risk.

Among the risks to be considered for a three phase separator : fire  and blocked outlet.

 

Our Three phase separator safety system is designed as follow:

In case of high pressure the PSHH will cause the SDVs to close. If the inlet SDV stack open, we are in the case of a blocked outlet, the PSV is designed to handle this scenario with a  relieving rate of 46000 kg/hr.

Regarding the fire scenario, whenever the fire is detected All SDVs are closed and the BDV will open.

So if the BDV stack closed we only have the PSV wich is designed for the blocked outlet case not fire.

 

Using Hysys simulation we can remark that the PSV is chattering during fire further more the fire scenario relieving rate is 3500 kg/hr.

 

Is this a safe situation !! 

 

So, if we want to convert the PSV to provide protection for fire scenario we will lack a safety layer for the blocked outlet scenario?

We are proposing that the PSHH of the three phase separator will directly shut-off the well head. But I really think that the production department will not accept such modification due to its high impact on production.

 

What should we do?

 

Best Regards,

 

Khaled

 

[fallah]

Khaled,

 

Please upload a simple sketch of the system you described...it will be more helpful than your description to submit a proper response to your query...

[kc2010]

Fallah,

 

Thanks for replying.

Please find attached a simple sketch of the system.

 

Regards

Attached Files

•  Sketch.pdf   969.64KB   95 downloads

[fallah]

Khaled,

 

A PSV can cover several scenarios, here fire and blocked outlet ones, but in your case the difference between two relief loads is so high such that for a conventional PSV chattering is much likely during the fire case relieving condition. Then, the matter could mostly be removed by using a modulating pilot operated PSV, even though in all fire cases it's mostly and really supposed that the PSV cannot protect the pressure vessel from failure and before the PSV reaches to relevant set point the vessel being ruptured due to much reduced allowable stress in its shell's metal especially when we see in your case you supposed the BDV being stuck in closed position...

[Bobby Strain]

No design engineer is concerned about relief valve damage resulting from a fire relief. Nor is such required by any code or recommended practice. So, don't stay awake nights worrying about it.

 

Bobby

[kc2010]

Fallah,

regarding the relieving temperature issue.

I've already used Hysys simulation to check the vessel wall temperature the results are shown in the file.

As you 've said the PSV will open after the design temeprature is reached.

 

However, I'm really not sure about the fiabilty of the result.

Do you think that the hysys simulation are given good results in this subject ?

 

The vessel material is  carbon steel with thickness of 16 mm and design temeprature of 120°C.

120°C is DT is given by process engineer I'm not sure if we have to use this value to judge that the PSV will not provide protection to the vessel during fire or it's another value related to mechanical properties of the vessel material !!

 

Further more if the PSV will not provide protection to the vessel during fire what should we do in order to be protected ?

 

Thanks.

Attached Files

•  Graphs.docx   29.7KB   31 downloads

[fallah]

Khaled,

 

The results can be reliable; because to specify the design temperature of a pressure vessel, the relieving temperature of the fire case isn't considered at all hence it's not strange if the PSV doesn't reach to the set point before vessel failure.

 

To mitigate the consequences of the fire case, even thouth a PSV being installed for code compliance, reliable protective measures such as depressurizing system, water spray and fire insulation can be applied...

[kc2010]

Fallah,

 

According To API 14C, We have to provide to provide two independent layers of protection for each identified risk.

The first one is the BDV system.

The second layer is PSV !! which is as we discused is useless .

Further more the water spray can't be considered as a second layer of protection as it's mainly made to provide assistance to the PSV function.

 

So, at this state we have an identified risk, Fire, with one protection layer, and we need to provide a second layer.

So what can we do !!

Do we have to made change !!

What kind of study we have to perform ?

[fallah]

Khaled,

 

In fact, depressuring system, water spray and fire insulation are protective measures just mitigate the consequences of the happened fire. Even a PSV which is installed to handle the fire case cannot mostly be considered as a protection layer due to lack of on time opening and vessel failure before PSV opening. Then, to meet the API 14C requirement for fire case appears, contrary to other scenarios in which the event happened firstly and then the protection layer will protect the system against overpressure , a protection layer has to be defined as a facility to remove the fire scenario among credible ones for the vessel by, let say, a fire wall or something like this... 

[Pronab]

Khaled,

Similar to your case we have three phase separator with PAHH(Pressure alarm HH) and it is an operator alert only. No signal is going to production well close. PCV & PSV are taking concern of HH pressure in the vessel.

 

Regards,

[kc2010]

Dear Expert Please let me summarize the discussion till Know, 

1) PSV is mainly made to provide overpressure protection whatever the scenario (Blocked outlet or fire)

2) In case of fire we can use: 

i) Two PSVs the first for the fire case (100 % MAWP) and the second (105 % MAWP) 

ii) Replace the Blocked outlet PSV by a SIL 3 PSHH (however we are actually using SIL 3 PSHH) 

iii) Using a Pilot PSV 

3) During Fire scenario PSV will not provide suitable protection due to induced stress failure.

4) BDV & water Spray are mitigation measure for the fire scenario.

So please let me ask a few questions:

1 ) Using rupture disk in case of fire as it will open at lower pressure due to heating , in fact this is stated by API-521, However my question : Does the rupture disk will open also during blocked outlet scenario? 

Or we can size it using a temperature set point i.e. it will open at 120°C regardless of pressure, and it will not open when the pressure reach 12 bara the PSV set point at normal operating temperature 90 °C ? 

2) API-521 Recommend not decreases in the environmental factor when using a water spray system. But if we suppose that the water spray system is highly reliable which value we can use for the environmental factor??

 3) Actually we are using a segregated blowdown strategy with two area production and GL, looking to the layout it seems to me they are in the same area furthermore Initially the platform was unmanned however it's not actually .

Shouldn't we use a total blowdown? 

Is there any industry standards or guidelines about this subject?

Attached Files

•  Layout.pdf   1.13MB   26 downloads

[Pan Nata]

 

Our Three phase separator safety system is designed as follow:

In case of high pressure the PSHH will cause the SDVs to close. If the inlet SDV stack open, we are in the case of a blocked outlet, the PSV is designed to handle this scenario with a  relieving rate of 46000 kg/hr.

Regarding the fire scenario, whenever the fire is detected All SDVs are closed and the BDV will open.

So if the BDV stack closed we only have the PSV wich is designed for the blocked outlet case not fire.

 

Using Hysys simulation we can remark that the PSV is chattering during fire further more the fire scenario relieving rate is 3500 kg/hr.

 

Is this a safe situation !! 

 

 

Best Regards,

 

Khaled

 

Just giving my two cents,

 

In my point view, there are two possibilities that we could be considered:

 

1.  As per your explanation, If CONFIRMED FIRE, I assume two SDVs which are installed in upstream and downstream of separator will get close to isolate the separator then BDV start to open with purpose to depressuring the isolated system. If BDV fail to open, PSV will guard the vessel from overpressure with Block Exit scenario since the protected system get isolated with SDVs because of fire.

 

2. PSV with fire case could be considered if fire happened then the two installed SDV being stuck open due to mechanical or instrument failure (system not get isolated) and those events occurred simultaneously. But IMHO, we could consider that situation as double jeopardy because those events (Fire and SDV failure) seems unrelated for me.

 

If you could provide us shutdown logic and/or cause and effects matrix maybe will be more clear for us.

 

However, as some expert here said, it is not easy to identify what is related and what is unrelated. Different guy maybe has different point of view and again this is just my tow cents

 

Rgds,