Cooling Water Bursting Disk 1

[aikmeam]

I've got a concern with a current design work I'm doing now which I hope to hear some opinions.

There are many water cooled heat exchangers in my work, in which the low pressure side (cooling water) pipings are not designed according to API 10/13 rule. And since the high pressure side can be as high as 1500# rating, we've installed bursting disk to prevent overpressure of the cooling water side in case of tube rupture in favour of its quick reaction time as compared to PSV. The setting of the bursting disk has been such where nuisance rupture can be avoided.

The tail piping of the rupture disk has been routed to flare as there might be a possibility of flammable gas from the high pressure side being relieved. My concerns are:

1. When the high pressure side finally loses its pressure after some time, cooling water will be continuosly drained into the flare, flooding the flare system. A relief scenario based on total plant loss of cooling water will then occur and since the header has already been flooded, there is a potential inability for the rest of the PSVs to vent to flare.

2. Is it really necessary to vent the gases + cooling water to flare? Thinking deeply, tube rupture is a remote case. A leak or pinhole is more likely, and that will not open the disk and eventually the flammable gases will be sent to the open atmosphere through the cooling tower anyway.

In the effort trying to be safe by venting to flare, it seems that like we're creating more hazard. I am thinking if venting it to safe location is a more pratical solution but then again, proving a safe location is indeed safe is not easy. Dispersion, radiation, and god knows what other thorough calculations needs to be done to prove that.

Appreciate anyone with prior experience on this same scenario to share. Thanks.

 

[EGT01]

Some time ago our company was evaluating a similar relief system for the cooling water side of an exchanger to protect against tube rupture. I wasn't directly involved but seem to remember the rupture disk discharged to some type of separator, maybe a sump that was vented. Unfortunately I don't have any more details to share.

You may want to review discussions in API RP-521 regarding
4.3 Atmospheric Discharge
4.5 Disposal to Lower-Pressure System
4.6 Disposal of Liquids and Condensable Vapors

Seems like you have given this some good thought. But here's a few more comments to consider.

Regarding your comment about flooding the flare system, seems that if you intend to vent the relief system to the flare you should ensure it is properly designed to accomodate the release.

Whether you vent to the flare or some other place, won't you have the potential to deplete your cooling water system if you don't react to the event to stop the water flow?

As the process side pressure falls, you will also have a possibility of getting water in on the process side.

I'm not sure what you are comparing in your second statement regarding venting of flammable gases at the cooling tower but atmospheric discharge of flammables gases is generally acceptable based on the practices in API RP-521. I don't think I would want to vent flammable gases at grade.

 

[Montemayor]

aikmeam:

What you are doing is already a latent improvement about to happen. The spirit of API 520/521 is to foster and lead safety in the work place and you are on the road there by studying the situation carefully and seeking advice where you can. I believe this Forum offers a lot of potential help in the area of expertise and experience in this area and I am looking forward to responses from others on this very important and practical engineering application. Allow me to address your concerns:

1) The flare system and its related KO pot (drum) need not be flooded in this scenario. Obviously, your flare system is an existing one, prior to the modifications you are proposing. As such, then, your present KO system at the flare is not designed nor scoped to handle the liquid load you are potentially thinking of imposing on it. You therefore must confront the fact that the existing KO system at the flare is going to be modified as part of your over-all program to protect the LP CWS. That having been said, lets agree that you have to include that in your scope of work. The correct size of KO equipment must be in place prior to installing the proposed CWS relief devices. There is no negotiating here, in my opinion. You’re going to have to do this whether you like it or not. Additionally, the concern for subsequent relief devices being triggered are not well-founded, I believe, because the scenario I envision is one of a plant shut down due to CWS failure. Therefore, all necessary Shut Down Valves (SDVs) should be activated by the incident of a tube failure in a key cooler-exchanger. This is not something I’m expounding; I am stressing the need to identify the criticality of such an incidence and the need to establish control over the total volume of CWS that would enter the KO drum – with the need to have other relief devices also activating. Certainly, the subject of a plant shut down is a serious event and requires careful study and hard decisions.

2. I believe it is really necessary to vent the gases + CWS to the flare via an appropriately-designed KO drum and related equipment IF the subject gases are flammable and/or toxic. This is the HEART of the matter: we’re talking safe emissions – not convenient “disposal”.

Of course, a leak or pinhole is more likely and gases could be channeled to the CW Tower and dispersed to the atmosphere there. BUT that is another, totally different case. For this eventuality, you should be monitoring and checking the CW Tower – especially if flammable and/or toxic gases are involved. The incident of your CW Tower catching fire is not beyond credibility here. Check out your insurance provider and you will find this out. Such a scenario will put you out of business, albeit temporarily. But it will result in a shutdown as well. You should be handling this problem simultaneously with instrument mitigation, in my opinion.

You are NOT creating another hazard by venting to flare. The flare site selection has been previously carefully chosen for safety and hazard reasons. Taking flammable and/or toxic gases to another location for “venting” disposal is not more practical nor safe. Stop and think about it: if it was, shouldn’t the flare be located at the second site? You’ll get logically killed by management if you offer this proposal as you stated it. Even if you go to a different, remote disposal site, you would still require an appropriate KO vessel and related equipment – and you’d still be facing a shutdown scenario!

Of course you’re going to have to re-visit the dispersion, radiation, etc. scenarios regarding your flare. But then, this is the normal, routine work involved with Management of Change (MOC) and would have to be done in any case under the Process Safety Management (PSM) principles. I’m assuming here that you and your organization are adhering to this under very strict guidelines. If you are not, then forget everything I’m recommending.

Your comment regarding nuisance ruptures related to bursting (rupture) disks is very interesting. This tendency of the disks to rupture prematurely due to corrosion, strain, age, calibration, defects, etc. is well known and documented. I would not, in our present age and engineering technology, ever employ and much less install another bursting disk again! We’ve come a long way from that technology with the advent of BUCKLING PINS. If you have not heard of this technology, I strongly suggest you find out about it ASAP. Even the manufacturers of bursting disks (BS&B, etc.) have seen the light and are offering buckling pins instead of burst disks. Buckling pins resolve most, if not all, of the above described problems with bursting disks. I wouldn’t install anything else but buckling pins for this application.

I hope these comments and experience helps you out. I envy your challenge and wish you luck and success. Make sure you document ALL calculations and decisions via PSM.


Art Montemayor
Spring, TX

 

[aikmeam]

Montemayor, a great piece of advice I must say.

Your comments on reworking the flare system is what I had in mind as well. But rather than modifying the whole flare system, I was thinking of adding a knock out pot at the outlet of each bursting disk for removal of the CW before sending the gases to flare. With this, I can say that the flare radiation and dispersion will not be affected.

Whether a plant shut down would occur before any other PSV starts popping cannot be easily predicted. It very well depends on the closing time of all SDVs to block out the whole plant. And PSV being the last level of protection, I would not want to risk it not being able to vent.

What I meant by safe location in my original post is a location high enough from the nearest men accessible platform. The idea came to my mind because flammable gases which I'm referring to are lighter than air and thus will not accumulate at low land. And coupled with cooling water which blows with it like a fountain, the chances of it being a real safety hazard is what I doubt.

I agree, safety should always be a priority concern. But in the end, there's always also the dollar sign to consider. And at times there's no need to be too conservative. It's always a balance between what hypothetically will happen and what actually happens.

Thanks for the recommendation of buckling pins. It's really new to me and I'll go through the web to see if I can find anything. Any recommended sites?

As for now, I'm still undecided. I have both options to weigh and I hope to hear more opinions and experiences from you engineers out there. Thanks.

 

[Montemayor]

aikmeam (Chemical):

Go to

http://www.rupturepin.com/

and also to

http://www.protectoseal.com/

for useful information on Buckling Pins. Rupture Pin Technology in Oklahoma told me that they have licensed the technology to ProtectoSeal, Inc. of Illinois, who is marketing and applying the buckling pins to industry. Protectoseal is an established, recognized and reliable manufacturer of conservation vents and other relief devices. They should be an excellent source of information, application recommendations and technical information.

Good Luck


Art Montemayor
Spring, TX

 

[fqureshi]

I have an experience on Ammonia unit where rupture discs were provided for the exchangers across the high pressure compressor inter and after coolers. The exhaust / flare system had a drip pot at the bottom with a U seal arrangement to separate and overflow water from the stream. A detailed and careful design of the bottom drip pot and U seal can eleviate your problem.

 

[aikmeam]

Interesting.. fqureshi. Almost similar situation as mine.

But I don't quite understand what u meant by "drip pot and U seal". Mind providing me a sketch?

Send to aikmeam@hotmail.com

Thanks.