Relief Valve Required?

[barnesed]

Steam header at 50 psig to the shell side of a heat exchanger with a design pressure of 150 psig. No other scenario for pressure in shell getting anywhere close to 150 psig. Is a relief valve required?

 

[Montemayor]

barnesed:

With no conceivable or credible source of pressure beyond 150 psig available to the shell side: yes you normally don't need a PSV to protect the shell side.

However, what is in the tube side? If you block in the tube side you could vaporize or expand the fluid in the tube side, rupture a tube(s) and over-pressure the shell side as well. Therefore, there is more to the story than what you've revealed. Until you've looked at the entire picture, you can't generalize.

 

[barnesed]

Thanks. The tube side is protected with a relief valve.

 

[jte]

barnesed-

You didn't state what codes are involved or where the exchanger is located. Assuming it is an ASME VIII-1 vessel in a typical US jurisdiction then I'll have to disagree with Montemayor on this one: You normally will have to protect the shell side.

Simply put, it is a requirement of VIII-1 to provide a relief device. The only exception is to use Code Case 2211 (Overpressure protection through system design). However, using the Code Case is something you must discuss (I'd suggest in writing!) with your Jurisdiction and insurance carrier.

See thread292-123430

jt

 

[BenThayer]

I agree with what Montemayor said (but who typically does not), there is more to the picture and you can't generalize.

You seem to have taken his response as an answer and I think he is really telling you take a more careful look.

Is the steam header truly protected at 50 psig or is it 50 psig because there is a regulator dropping the primary header pressure from 175 psig? A "failed" regulator scenario would require a PSV somewhere. Either on the header or on the exchanger.

Regarding the tubeside, so what if it has a relief. For an exchanger that has a rating of 3000 psig on the tube side with a 3000 psig relief, a failed 3/4" tube could reek havoc with a 2500 psig process in the tubes.

You should review some of the API scenarios for relief and then start making some decisions of if you need a relief valve.

 

[barnesed]

Sorry, in trying to be brief, I intentionally left out some details. The steam is generated at 105 psig with relief at 125 psig. After reducing to 50 psig, there is another relief valve. Therefore, the steam to the exchanger will never be close to 150 psig. The tube side is also rated at 150 psig with water flowing at ~50 psig. The pump supplying the water is not capable of generating a pressure of 150 psig, even if deadheaded. The tube side relief is for thermal expansion and is set at 150 psig. Without a credible shell side overpressure scenario, I was fishing to see if anyone raised a code issue. The exchanger is 9'-6" in length with an OD of 1'-6".

 

[TBP]

If the steam valve fails to shut, combined with the water side being bottled-up, either because it's a closed loop, or isolation valves are closed, you can get some pretty impressive hydaulic pressures. You can generate some amazing pressures just having a pipe or vessel filled with cold water left bottled-up overnight in hot weather. Personally, I'd install a relief valve on the water side. It's pretty cheap insurance.

 

[jte]

Section VIII doesn't care much about whether or not your overpressure scenario is credible or incredible. Only Code Case 2211 gets into that. With a straight VIII-1 vessel (is it VIII-1?) you have a code issue with part UG-125 which mandates a relief device irrespective of HAZOP scenarios.

You indicated that after reducing the pressure of the steam header to the shell side to 50 psi you have another relief valve set, presumably, below the 150 psi MAWP of the shell side. Provided you have only CSO full port block valves between the relief valve and the shell, along with adequately sized piping, it appears that you can meet the requirements of the code as long as the existing valve on the piping has adequate capacity.

jt

 

[MJCronin]

Barnes, Montemayor, et al....

I agree with the all of the statements noted above..... however, I think it is important to recognize the fragmented nature of the codes, standards and rules that chemical engineers have to deal with.

The HX shell has been designed, certified and tested under ASME VIII. A relief device is absolutely required for the shellside under ASME rules.

The steam piping system requires a relief device at the pressure reducing staion.....under the rules of the ASME B31.1./B31.3 Piping Codes

The tubeside of the HX requires thermal protection against tube rupture based on a OSHA mandated HAZOP analysis. The rules and "recommended practices" of API (American Petroleum Institute) are commonly used for this kind of evaluation.

In the USA, it is this fragmented nature of codes, standards, guidelines and practices which is the source of much confusion and of many problems that should never occur.

How many shell and tube HXs are there out there with exactly the same configuration ?.....Thousands ???? Why should there be so many expensive guidelines that must be consulted befor an adequate system can be designed ??

If the AIChE and/or the US Chemical Safety Board had any real "testicular fortitude" and had the will and resources to exert any real authority, there would not be so many interlocking and sometimes conflicting standards and guidelines.

My opinion only.........

MJC

 

[TrevorP]

It seems from what you've said that you already have pressure relief on both shell and tube sides, protected from the only available sources of overpressure. Without knowing the full details, it seems that you do therefore comply with the code.
Is the unit rated for full vac? You could get full vac generated on the shell side by steam condensation.

 

[PAN]

barnesed,
This may be off the topic. You should mention about temperature for both shell and tube side. Is it possible for your service to vaporize water in the tube side?

 

[JimCasey]

Maybe I am overly simplistic, but you have a code vessel and if there is EVER a problem with that vessel and there is not a code SRV on it, you really won't enjoy going through the inquisition. You might have it pencil-whipped where you don't need a SRV NOW, but what happens if an adjacent system is revised and that affects this vessel?

 

[Montemayor]

All:

In my senility years, I may be missing the part where barnesed, the original poster, states that his exchanger is designed and stamped with the ASME Section VIII stamp. But I don’t think so, because my grandchildren have looked at it and they don’t see where the word “stamp” is hidden. Or maybe one of barnesed’s latest postings was deleted. I don’t think so either, because I’ve been following this thread.

Just about every exchanger I’ve designed, built, operated, and maintained has been designed using ASME Section VIII. There are some, like the Graham Heliflow and others that don’t adapt to some of the criteria in Section VIII. Those are exceptions. And just about everyone knows that Section VIII mandates that a relief valve shall protect a pressure vessel. And that, in my opinion, is wise engineering. Heat exchangers and other pressure-contained equipment can be designed by the Section VIII code; but they don’t necessarily have to be stamped as such. There are such things as non-stamp states in the USA and non-stamp countries. I don’t know which state or country barnesed’s exchanger is located; he hasn’t stated that. Nor has he stated that the exchanger is stamped. Careful reading of his posts will confirm that fact.

My point is: just because an exchanger is designed as per Section VIII doesn’t mean that it HAS to have a PSV. However, if the same exchanger is designed and stamped with Section VIII and local codes and regulations mandate adherence to ASME’s rulings, then by all means put a PSV on it. That’s the simple answer. However, unless someone shows my feeble eyes where in barnesed’s 3 postings he has stated that his exchanger is stamped and is under the jurisdiction of Section VIII via the local codes, I can’t just assume that. That’s why evaluating the process and its potential hazards is the absolute and correct engineering manner to approach such a situation. Just applying a PSV because “ASME told me so” is not properly designing a process vessel or other equipment. I consider such an attitude an engineering “cop-out”. I believe I understand the spirit under which ASME mandates a PSV – and it isn’t done to replace or substitute for the local engineering safety analysis of the application.

I heartily agree with MJCronin’s truthful and frank comment regarding the fragmented nature of the codes, standards, and rules that all engineers have to deal with. But we have, in the ultimate end, the local codes and regulations fixed on the operating equipment that dictate the minimum safety criteria to be legally installed. If the local codes and regulations fail to protect the operator and the public, we have only ourselves to blame. Insurance companies can help; but they are in the game for profit, not for protection of the public. Again, I emphasize the detailed process analysis of the safety hazards involved. Nothing that I’ve come across in the past can specifically improve the safety as can this method – and this includes Hazop, safety meetings, PSSR, MOC, etc.

 

[jte]

Montemayor-

I'll point out that I stated my assumption in my post of 6 Jul 05 that the exchanger was a VIII-1 in a typical US jurisdiction. Yes, there are some non-code states, but the large majority of US states are code states.

In his response of 6 July 05, barnsed does not dispute my VIII-1 assumption, and provides dimensions which along with the design pressure would put the exchanger squarely into the scope of VIII. Further, he states in his post that "...I was fishing to see if anyone raised a code issue."

OTOH, you did not state your assumption that the exchanger was located in a non-code state or country 'till your post of 15 July 05.

So... It would seem that unless and until barnsed returns to this thread, it is not unreasonable to presume that in fact, the subject exchanger is stamped to VIII-1 and is in a jurisdiction which cares.

By the way, I appreciate that you are bringing your grandkids into the wonderfully fuzzy and gray world of engineering! Perhaps one day, a few decades from now, I'll be getting my grandkids involved also!

jt