6 replies to this topic

[rikakose]

The reactor is equipped with full length cooling jacket. 

Cooling water is always on. Outer shell is stainless steel. 

 

Can we take water jacket as reduction factor in the fire case?

 

Cooling water is 6 barg, 10m3/h

Wet surface in the reactor is around 12 m2. 

 

BP of the solvent in the reactor at relief pressure (5barg) is 120ºC

 

The cooling water can absorb actually all the energy from fire. 

 

How to determine the relief load of the PSV?

 

Should it be considered as steam control failure case? Then the heat transfer capacity is limited by the reactor wall. 

Edited by rikakose, 28 August 2019 - 07:01 AM.

[PhilippM]

You should probably discuss this during a HAZOP and decide if the cooling water supply is regarded as sufficiently reliable to cool the reactor during an external fire event. I would probably just assume that all flows (including cooling water) are stopped in case of a fire and not take the cooling water into account at all.

 

The jacket might be taken into account by using a smaller F-factor than 1, some clients have standards that let you do that.

The calculation to determine the relief load is described in API 521.

[rikakose]

You should probably discuss this during a HAZOP and decide if the cooling water supply is regarded as sufficiently reliable to cool the reactor during an external fire event. I would probably just assume that all flows (including cooling water) are stopped in case of a fire and not take the cooling water into account at all.

 

The jacket might be taken into account by using a smaller F-factor than 1, some clients have standards that let you do that.

The calculation to determine the relief load is described in API 521.

Thanks for the reply.

Fire case was not included in the PSV calculation. Now a new product is introduced, in the new HAZoP, they take fire case as a valid case. 

Current relief system is way too small for the fire case (without taking into account the mitigation of cooling water) . New product cannot wait for the modification of the relief system.

 

Or, is there a way that i can exclude the fire case?  

[latexman]

Describe the construction detail of the jacket please.  If there is metal that contacts or connects the outer shell to the inner shell, like a half-pipe or full jacket with spiral baffles, I doubt you get much credit

[rikakose]

It is simple annular jacket. see the sketch. 

 

Attached Files

•  sk3.pdf   147.13KB   18 downloads

Edited by rikakose, 30 August 2019 - 08:19 AM.

[rikakose]

The floor is sloped under reactor. The chance of having pool of hydrocarbon is low. I think the fire case can be excluded. 

Then the client safety engineer comes up with the jet fire idea. His argument is: the reaction pressure of the new product is 4 barg (reactor design 5 barg). Old product produced at 1 barg. So the chance of leaking is higher........

[latexman]

You should check what your company allows.  This is what MY company allows:

 

Vessel Jackets

 

Credit for jackets on vessels can only be taken when the jacket design is an unbaffled outer shell

over the vessel similar to the German Din 28136 design. Examples of this design include

Pfaudler and DeDietrich glass-lined vessels with standard jackets. No credit is allowed for

baffled jackets, dimpled jackets, pipe coils, or similar higher heat transfer jackets.

For vessels with annular jackets meeting the above criteria, the following fire credit factor can be

applied:

1. Calculate the vessel effective area as “Wetted area not under jacket + 0.7 * wetted area under

jacket”.

2. Insulation factor of 0.3 may be applied to the effective area, if properly insulated.