RAGAGEP for fire case PSV's on shell and tube heat exchangers

[sanderson231]

I retired last year after spending 41 years in the oil refining business. The last 36 with a large independent refiner. I have been doing some post retirement consulting work. OSHA now requires equipment to be designed using Recognized And Generally Accepted Good Engineering Practice. Where I came from, the company had been installing fire case PSV's on the shell side of heat exchangers for many years. Recently the company began installing fire case PSV's on the tube side of heat exchnagers. I recently became aware of two large operating companies that do not consider it necessary to install fire case PSV's on either side of a heat exchanger. I talked to two large engineering, procurement and construction companies and they consider RAGAGEP to be installing PSV's on both shell and tube side. Is there an industry concensus on what is RAGAGEP? From my perspective both the sheel and tube side are pressure vessels and it would seem like they should have fire case protection of some sort. Either a PSV or a guaranteed open path to another vessel which does have a large enough PSV.

 

[EmmanuelTop]

Fire protection for tubes - that seems like a great overkill and definitely not something coming from a "RAGAGEP" approach. Detailed engineering evaluation would probably eliminate fire PSV for exchanger tubes in 90%+ cases.

The way I understand the definition of good engineering practice is: to always be 100% sure why something is installed, or why something does not need to be installed - not that it has to be installed in 100% cases. Then it becomes a code, and it is not an engineering practice anymore.

Every prescriptive design misses the point, more or less, and the same applies with fire PSV allocation.

Dejan IVANOVIC
Process Engineer, MSChE

 

[don1980]

The sizing basis for pressure vessel relief devices is properly left to the judgment of the equipment owner. Thus the owner is free to decide whether an exchanger needs a fire-sized relief device. My observation is that exchangers in the refinery industry are generally not equipped with relief devices sized for fire exposure. If the flow paths through the exchanger are open during a fire, then I think it's perfectly reasonable to not install a fire-sized relief device on the exchanger.

 

[georgeverghese]

Adding to responses, some brief rules on firecase relief on S/T heat exchangers are ( which are abstracted and summarised from company guidelines which I hope wont be considered a breach of confidentiality) :
a) Firecase relief is not applicable on a HX side if this occurs as a result of manual isolation block valves in the closed position for maintenance only. These valves should be locked open and only operated under regulated work permit conditions after complete draining of inventory.
b) If the HX can get blocked in by some means during operation other than (a) ( for example a closed control valve on one side and a closed shutdown valve or check valve upstream), then firecase thermal relief provisions need only be provided on the shellside, provided shellside inventory is in excess of 500litres. Tubeside relief for this case is typically assumed to be adequately accomodated at the tubeside channel flanges. Exceptions to this are when (a) shellside volume is less than 500litres but inventory is toxic or is TEMA type N on one end and type L, M or N or the other end (b) tubeside configuration does not have body flanges or is some special high pressure closure design.

 

[EmmanuelTop]

This article might be an interesting read: Understanding Overpressure Scenarios and RAGAGEP


Dejan IVANOVIC
Process Engineer, MSChE

 

[georgeverghese]

Omission in previous response:

Tubeside firecase vapor relief for this case is typically assumed to be adequately accomodated at the tubeside channel flanges. However, provisions for liquid hydrostatic overpressure relief on the tubeside are required if the tubeside is predominantly liquid full and the tubeside could be blocked in.

 

[sanderson231]

Thanks for all the responses so far. It does seem that there is far from a consensus in industry. However the reason I am asking about RAGAGEP is that it is now an OSHA requirement.

https://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=INTERPRETATIONS&p_id=30785

OSHA considers consensus codes such as ASME boiler/pressure vessel codes and API recommended practices such as RP-520 and RP-521 to be RAGAGEP. In the past following API recommended practices may have been optional but with OSHA's current stance it is seems like there is now a legal requirement to follow the code and recommended practices.

Heat exchangers are typically designed using the ASME Section VIII pressure vessel code. The name plate on the exchanger has a "U" stamp with both shell side and tube side MAWP's. Therefore the channel, shell, and back bell of a heat exchanger are pressure vessels. To satisfy ASME code and API recommended practices, fire case over-pressure needs to be looked at for pressure vessels.

So to meet OSHA requirements it seems like there are three possibilities for liquid containing heat exchangers.

1) Install PSV's on both the tube side and shell side
2) Guarantee an open path with acceptable pressure drop to another PSV which has a large enough PSV
3) Assume that the studs on the heat exchanger relax in a fire and fluid leaks out at a fast enough rate to keep
the heat exchanger from over-pressuring and rupturing.

Option 1) clearly meets ASME code and ASME RP's. It just costs a bunch of money.

Option 2) is easy in some case e.g. a reboiler connected to a distillation column with a large PSV with block for on-line maintenance. Just car seal or lock the valves open. It is difficult when there is intervening equipment e.g. additional heat exchangers where it is difficult to predict pressure drop or when there are control valves in the system ( can we be sure of the position of a control valve in a fire?)

Option 3) has logic to it. Many times it can be shown that a vapor filled vessel will structurally fail in a fire due to wall overheating prior to reaching it MAWP. Therefore fire does not need to be considered as an overpressure scenario when sizing a PSV. But can we be sure the studs on a heat exchanger relax in a fire to make it self relieving? Are there industry examples of this happening? Are there industry examples of this not happening?

 

[georgeverghese]

For option 2, even a rough estimate of pressure drop should do. For intervening control valves in the relief path, we must assume that these may be either fail closed or forced close by the control system during a relief event.

For option 3, in addition to the body flange leaks, there will also be the tubeside liquid hydrostatic thermal RV. Moreover, most of the heat absorbed on the tubeside channels would get dissipated through the tubes into the shellside relief device. These are company guidelines from industry experts so I can only presume these are well founded guidelines. Pls also note exceptions to this in previous response.

 

[don1980]

Sanderson - I think you're assuming that ASME and API have prescriptive requirement which mandate sizing for fire exposure. That's not true. ASME and API tell you what to do after you have decided to size for fire exposure. All sizing basis decisions, including fire, are intentionally left to the user. Regarding the fire decision, it's up to the user to determine combustibility criteria (what fluids are a fire risk) and it's up to the user to assess when fire exposure protection is needed.