Credit for sprinker systems in RV sizing. 1

[jamesbanda]

Dear Sirs,

I've been reviewing some "very old" relief calculations from a now redundent plant - because they are comparable to an exsiting design...

On those calculations credit is taken for sprinker protection to reduce the fire case by 75%. Due to the toxic nature of the chemicals using large relief valves itself poses a safety consern becasue the relief of those to atomsphere would be consern.. So my question has anyone else seen this... its been hiNted that this is an old UK code from ICI..

I understand API521 Is from refinarys and does not allow credit but this is becuase many relinary installations are remove and deluge quality/reliablity may not always be high (ie ia desert..).. IS THIS TRUE ??


Note: Our chemicals cannot be flared due to issues with combustion products..

SO..HAS ANYONE SEE THE USE OF THE CODE LIKE THIS??

 

[MortenA]

Sizing for the fire case in accordance with API shall not consider deluge.

The reason for this is that your deluge shall make sure that the PSV is NEVER used!

The PSV is there if your deluge fails!

Best regards

Morten

 

[JoeWong88]

Jamesbanda,

When we are dealing with safety, two aspects shall be in mind.
1. Safe design
2. Code

Safe design
"...credit is taken for sprinker protection to reduce the fire case by 75%...". Is this Safe design ? In the context, how would ensure availability, reliability and effectiveness of sprinkler service ? If not, should the credit to be taken ?
If yes, why not 100% ? How to justify 25% credit ? Is there any Quantitative Risk Analysis being conduct to justify the 25% credit ? If you don't have the answer, then you are not in the position to take the credit.

code
You have to check the basis document which code you shall comply to. It could be API, JIS, GB, DIN, etc. Each code may have different consideration.

You also have to check the local regulation requirement. In few events per my experience, the local code is even more stringent than API.

Check it out.

 

[jamesbanda]

Thanks for your comments..

However, I am not sure why this case was taken as 75%. A reputable EPC firm selected the case based on old design codes.. .

Personnlly I think the sizing case was questionable.. i fail to see why a reputable EPC firm would make this call either.. however i want to see if any experts out in industry have ever seen this..

Having very large relief valves are not neccessarily safer than small ones... large relief valves on some systems (if accidentially they lift) will result in large toxic releases..

 

[MortenA]

But an undersized RV may cause the vessel to explode.

Best regards

Morten

 

[JoeWong88]

Jamesbanda,

I have nothing much to inject into your query...

But i love to know and learn the reference for "old design code". Probably some of experience engineers can recall...

I dare not to say NO to 25% credit but i also dare not to ACCEPT 25% credit without any JUSTIFICATION & COMPLIANCE with LATEST CODE and REGULATIONS...If i have no choice, i will take the conservative approach.

The reputable EPC company dare to take 25% credit and managed to convince client and regulator to accept it as this EPC company has conducted series of activities e.g. risk based analysis, QRA, SIL studies, Dynamic sturdies, etc to check and confirm it. Based on conclusion of these studies, the engineers raised DEVIATION with these supports. DEVIATION is common but you need a strong SUPPORT.

API stressed that FIRE case is one of the case shall be addressed during design...however...I have even seen people eliminate FIRE case...Extra Caution When Eliminating Overpressure by Fire Attacks

A simple point...DO NOT provide OVERSIZED or UNDERSIZED PSV...OVERSIZED lead to potential PSV chattering and excessive relief load...UNDERSIZED lead to potential Catastrophe...

JoeWong
Chemical & Process Technology

 

[unclesyd]

I just checked with our relief valve expert (30 yrs) about taking credit for deluge systems in sizing PSV/PRVs. His answer was that he has never taken credit nor does our corporate standards allow any credit for deluge systems. He also said that at a couple of seminars he has heard and understood that some sites do take credit. Again he said hasn't seen anything in writing.

Our site has an extremely large deluge system in the process area where we air oxidize cyclohexane. There are two trains each with 6 ea 100,000 gal reactors that are wrapped in sprinkler piping but no credit is given on relief valve sizing. There are 2 12" PSVs on each reactor. There are also several PRVs set to relieve to a remote dump tank in case of fire.

We also have a Therminol vaporizing system with area consisting of 5 D tube vaporizers in close proximity to each other with an enclosed control room. This area is protected with a extensive deluge system. Each vaporizer has an 8" PSV on the vapor drum which is vented to the atmosphere. An event in 1988 caused two tubes in one of the vaporizers to rupture at the same time overwhelming the sand trap designed to capture the flow of a single tube. This event caused the fire box to be pressurized and in turn caused the control room to be abandoned and also prevented access to the manual block valves on the feed to the vaporizers. At this time the system was essentially venting all the Therminol system, both vapor and condensate, through the ruptured tubes. The non-return/block valve on vapor drum didn't seat as the system pressure stayed up until the remaining vaporizers could be shutdown. When the non-return valves closed it still allowed the condensate, feed, to feed the affected vaporizer. There was an explosion vent on top of the vaporizer that partially opened which caused a tongue of flame to be directed toward the PSV on vapor drum. . The tongue of flame that was hitting the PSV eventually caused the valve to lift due to a weakened spring. The tongue of flame from the PSV vent pipe caused the flame to impinge on the stack for this vaporizer to the extent that it started bend at the 20' level to approximately 45 degrees before water was directed at it. The fire was never extinguished it went out when it ran out of fuel.
We had become complacent with this system over the years since it always operated without any problems for 35 years.
The remedial action was to vent the PSV to the stacks and install drains for same. The deluge system was modified to cover the PSVs to prevent a repeat of this type event. The feed system piping was modified by adding automatic valves in the feed line at each vaporizer. Additional manual valves were added away from the area. Crash buttons to shutoff and dump the instrument air were added at the control room. When the control system was modernized access was provided from remote panels. In addition the PRV/PSVs in the cyan area were provided with additional sprinkler heads.

This system operates at 70 psig and we only use the latent heat of the Therminol so the vapor and condensate are essentially the same temperature.

The deluge system worked quite well in protecting the other equipment. The plastic lamp shades around the enclosed light bulbs were not touched by the fire.

Even though this system was thought to be quite reduntant we overlooked one thing, you have to protect the valves that do the protecting.